AnthropicLanguage modelActive
Claude Fable 5.2
Claude Fable 5.2 is among the strongest models: Rank 1 of 94 in the overall index. First place is not certain (win chance 73%): Gemini 4 Carbon and Claude Opus 5.5 are statistically level. Through the API it is expensive.
IntelligenceRank 1 of 94 · top group
130.9Overall indexhigher is better
Speednot measured
n/ano measurement yet
PriceRank 97 of 100
Input $8 · Output $40
$16per 1M tokenslower is better
ContextRank 36 of 110
1Mtokens at once≈ 1,500 A4 pages
Compare with another model
Summary
It is expensive: $16 per 1M tokens (input and output blended); a typical model costs $1.50.
Claude Fable 5.2 became available on November 9, 2026, succeeding Claude Fable 5.1. Anthropic positions it for long agent sessions, demanding coding tasks, and multistep research. It accepts text and images and produces text.
Standard API pricing is $8 input and $40 output per million tokens, 20% below its predecessor. Adaptive thinking cannot be disabled. Effort levels from low to max control the work performed, with high as the API default. Anthropic recommends starting with Claude Opus 5.5 for most tasks.
Prices
5 offers, prices in US dollars per 1M tokens, blended from 3 parts input and 1 part output.
| Provider | Input | Output | Cache read | Cache write | Blended |
|---|---|---|---|---|---|
| Anthropicdirect | $8 | $40 | $0.20 | $10 | $16 |
| Amazon Bedrock | $8 | $40 | n/a | n/a | $16 |
| Azure | $8 | $40 | n/a | n/a | $16 |
| GoogleGlobal | $8 | $40 | n/a | n/a | $16 |
| AnthropicFast | $16 | $80 | n/a | n/a | $32 |
Cache prices come from the direct tariff; OpenRouter does not report them for the other providers. Direct price according to the developer, checked Nov 9, 2026. Provider prices via OpenRouter, as of Nov 11, 2026. Open the LLM price index
Unchanged since tracking began on Nov 10, 2026.
InputOutput
At Anthropic from $20 a month (Claude Pro). Through other providers from $39 a month (GitHub Copilot Pro+, with limited credits). 10 plans in Germany.
| Plan | Monthly list price | Note |
|---|---|---|
| Claude ProAnthropic | $20 | developer's own plan |
| Claude Team Standard-PlatzAnthropic | $25 | developer's own plan, with limited credit, for teams, per seat |
| Claude MaxAnthropic | $100 | developer's own plan, with limited credit |
| Claude Team Premium-PlatzAnthropic | $125 | developer's own plan, with limited credit, for teams, per seat |
| Claude UltraAnthropic | $200 | developer's own plan, with limited credit |
| GitHub Copilot Business | $19 | with limited credit, for teams, per seat |
| GitHub Copilot Enterprise | $39 | with limited credit, for enterprises, per seat |
| GitHub Copilot Pro+ | $39 | with limited credit |
| GitHub Copilot Max | $100 | with limited credit |
| Perplexity Max | $200 | with limited credit |
Claude Fable 5.2 benchmarks
Overall index (Gradually Meta-Score v5)
20 of 94 models
4 models in this view are measured at a lower reasoning effort than max in the overall index. Dashed column: lower reasoning effort than the top.
Claude Fable 5.2max130.9
Gemini 4 Carbonhigh127.0
Claude Opus 5.5max125.6
Claude Fable 5.1max124.1
Claude Sonnet 5.5max121.8
Gemini 4 Argonhigh121.4
GPT-6 Astramax120.2
Muse Wonder Fullmax118.1
GPT-6.1 Solmax118.0
Claude Opus 5max117.4
Muse Wondermax113.3
Claude Fable 5max112.8
Kimi K3.1 Promax112.6
Muse Ember 1.0max110.4
GPT-6 Solmax109.1
GPT-5.6 Solmax109.1
Gemini 3.7 Flashhigh107.9
Muse Spark 1.3max107.3
Gemini 3.8 Flashhigh106.8
DeepSeek V4.1 Flash 1017max106.7
Claude Fable 5.2 by reasoning effort
Points on the overall index scale. The value at max is Claude Fable 5.2's overall index; the other efforts derive from the same Artificial Analysis test series. Missing efforts are never estimated.
- low121.895% range 111.9 to 131.7
- medium123.895% range 114.4 to 133.1
- high127.295% range 118.6 to 135.7
- xhigh128.095% range 119.6 to 136.5
- maxeffort of the overall index130.995% range 122.6 to 139.2
Source: Meta-Score v5; tiers: Artificial Analysis
Intelligence vs. price
most attractive area: stronger and cheaper than half of the models shownbest intelligence for the price
More models in the chart (27)
- Claude Fable 5.1124.1 · $20
- Gemini 4 Argon121.4 · $4
- GPT-6 Astra120.2 · $20
- GPT-6.1 Sol118.0 · $4
- Claude Opus 5117.4 · $10
- Muse Wonder113.3 · $8
- Claude Fable 5112.8 · $20
- Kimi K3.1 Pro112.6 · $4
- Muse Ember 1.0110.4 · $0.20
- GPT-6 Sol109.1 · $4
- GPT-5.6 Sol109.1 · $8
- Muse Spark 1.3107.3 · $2
- Gemini 3.8 Flash106.8 · $1.50
- DeepSeek V4.1 Flash 1017106.7 · $0.26
- GPT-6.1 Terra106.5 · $2
- Grok 4.7103.3 · $3
- Claude Haiku 5.5102.8 · $0.20
- Grok 4.6102.1 · $3
- GPT-5.5101.5 · $11.25
- Kimi K3101.1 · $6
- GPT-5.6 Terra100.8 · $4.50
- Mistral Large 4100.8 · $1.05
- Qwen 3.8 Max 0902100.8 · $2.48
- GLM-5.3100.6 · $2.15
- MiMo V2.6 Pro100.4 · $0.54
- MiniMax M3.1 Flash100.0 · $0.26
- Claude Opus 4.899.5 · $10
Claude Fable 5.2 by task area
Points, 100 = reference group average · higher is better
Knowledgetoo little data145.5
Mathematicstoo little data135.2
GeneralRank 1 of 69132.8
CodingRank 1 of 114131.6
ReasoningRank 2 of 76124.9
AgentsRank 1 of 100124.2
FinanceRank 2 of 113123.5
LegalRank 4 of 81120.9
Light bar: too few results for a rank yet. No tests yet: Multimodal, Health.
Best LLM by task:CodingReasoningMathKnowledgeAI agentsGeneral tasksFinanceLegal
Terminal-Bench 4.0
Share of tasks the model solves as an agent in the terminal, in percent · higher is better
20 of 40 models
Claude Fable 5.265.2 %
Claude Opus 5.565.2 %
Claude Sonnet 5.564.1 %
Kimi K3.1 Pro63.6 %
Gemini 4 Carbon63.1 %
GPT-6 Astra59.6 %
Claude Fable 5.158.1 %
Muse Wonder Full58.1 %
Gemini 4 Argon57.6 %
GPT-6.1 Sol55.1 %
Claude Opus 553.5 %
GPT-6.1 Terra49.5 %
Muse Wonder49.5 %
GPT-6 Sol44.4 %
Claude Fable 541.4 %
GLM-5.338.9 %
GPT-5.6 Sol37.9 %
GLM-5.3-Flash37.4 %
Claude Haiku 5.536.4 %
Qwen 3.8 Max 090234.3 %
as of November 10, 2026 · Source: Vals AI
Vibe Code Bench
Share of apps the model builds correctly from a description, in percent · higher is better
20 of 53 models
Claude Fable 5.293.8 %
Claude Sonnet 5.592.4 %
Gemini 4 Argon91.9 %
Claude Haiku 5.590.4 %
Claude Fable 590.4 %
Claude Opus 5.590.3 %
Claude Fable 5.190.3 %
GPT-6 Astra89.6 %
Gemini 4 Carbon89.1 %
GPT-6.1 Sol88.9 %
Claude Opus 588.4 %
GPT-6 Sol87.8 %
DeepSeek V4.1 Flash 101787.7 %
Grok 4.786.2 %
Muse Wonder Full86.0 %
Muse Spark 1.385.9 %
Kimi K385.0 %
Kimi K3.1 Pro84.9 %
DeepSeek-V4.1-Flash84.7 %
GPT-6.1 Terra84.0 %
as of November 10, 2026 · Source: Vals AI
Show all 289 published benchmark results
Page 1 of 25
Measurements without matching peer values
These measurements have no matching peer values under the same test conditions. Their original values and sources remain available here.
AutomationBench-AA: Overall (Show measurement, test conditions, and source)
- Source value
- 68.82266338432704
- Metric
- Objectives completed without guardrail violations
- Unit
- %
- Benchmark version
- AA private 657-task split of AutomationBench v1.0.6
- Category
- agentic
- Task
- overall
- Task label
- Overall
- Direction
- Higher is better
- Comparison cohort
- artificial-analysis-automationbench-aa:AA private 657-task split of AutomationBench v1.0.6:v5.0.1:reasoning:effort=medium:suffix=Medium%2C%20Default%20Fallback
- Source type
- independent-evaluator
- Evaluator
- Artificial Analysis
- Status
- active
- Retrieved at
- 2026-11-09
- Methodology
- Exact AA configuration: Claude Fable 5.2 (Medium, Default Fallback) (release claude-fable-5-2, row claude-fable-5-2-medium, UUID 0c9b4e05-9259-42cc-b40e-d6acbbeaa7e8, reasoning=true, effort=medium, suffix=Medium, Default Fallback). This source capture does not provide an average reasoning-token count.. Source model ID: artificial-analysis/0c9b4e05-9259-42cc-b40e-d6acbbeaa7e8/claude-fable-5-2/claude-fable-5-2-medium. Source model name: Claude Fable 5.2. Provider reported by source: Anthropic. Model snapshot: Claude Fable 5.2 (Medium, Default Fallback). Evaluator: Artificial Analysis. Dataset type: artificial-analysis-evaluation. Reasoning: reasoning enabled. Reasoning effort: medium. Score display: 68.8%. Source ID: artificial-analysis. Evidence capture ID: aa-model-page-2026-11-09-7542f902e8e4. Source payload hash: 7542f902e8e4f947d83dc15004997367540a045beb71bef86f98e77adbc6a23a. Parser version: artificial-analysis-rsc-models-v4. Imported at: 2026-11-09
- Comparison methodology
- Exact AA configuration: Claude Fable 5.2 (Medium, Default Fallback) (release claude-fable-5-2, row claude-fable-5-2-medium, UUID 0c9b4e05-9259-42cc-b40e-d6acbbeaa7e8, reasoning=true, effort=medium, suffix=Medium, Default Fallback). This source capture does not provide an average reasoning-token count.. Reasoning: reasoning enabled. Reasoning effort: medium
AutomationBench-AA: Overall (Show measurement, test conditions, and source)
- Source value
- 76.42233832611831
- Metric
- Objectives completed without guardrail violations
- Unit
- %
- Benchmark version
- AA private 657-task split of AutomationBench v1.0.6
- Category
- agentic
- Task
- overall
- Task label
- Overall
- Direction
- Higher is better
- Comparison cohort
- artificial-analysis-automationbench-aa:AA private 657-task split of AutomationBench v1.0.6:v5.0.1:reasoning:effort=max:suffix=Max%2C%20Default%20Fallback
- Source type
- independent-evaluator
- Evaluator
- Artificial Analysis
- Status
- active
- Retrieved at
- 2026-11-09
- Methodology
- Exact AA configuration: Claude Fable 5.2 (Max, Default Fallback) (release claude-fable-5-2, row claude-fable-5-2, UUID 593f2d23-247b-41c9-a679-cebd225f0a9a, reasoning=true, effort=max, suffix=Max, Default Fallback). This source capture does not provide an average reasoning-token count.. Source model ID: artificial-analysis/593f2d23-247b-41c9-a679-cebd225f0a9a/claude-fable-5-2/claude-fable-5-2. Source model name: Claude Fable 5.2. Provider reported by source: Anthropic. Model snapshot: Claude Fable 5.2 (Max, Default Fallback). Evaluator: Artificial Analysis. Dataset type: artificial-analysis-evaluation. Reasoning: reasoning enabled. Reasoning effort: max. Score display: 76.4%. Source ID: artificial-analysis. Evidence capture ID: aa-model-page-2026-11-09-7542f902e8e4. Source payload hash: 7542f902e8e4f947d83dc15004997367540a045beb71bef86f98e77adbc6a23a. Parser version: artificial-analysis-rsc-models-v4. Imported at: 2026-11-09
- Comparison methodology
- Exact AA configuration: Claude Fable 5.2 (Max, Default Fallback) (release claude-fable-5-2, row claude-fable-5-2, UUID 593f2d23-247b-41c9-a679-cebd225f0a9a, reasoning=true, effort=max, suffix=Max, Default Fallback). This source capture does not provide an average reasoning-token count.. Reasoning: reasoning enabled. Reasoning effort: max
AutomationBench-AA: Overall (Show measurement, test conditions, and source)
- Source value
- 74.16293165236524
- Metric
- Objectives completed without guardrail violations
- Unit
- %
- Benchmark version
- AA private 657-task split of AutomationBench v1.0.6
- Category
- agentic
- Task
- overall
- Task label
- Overall
- Direction
- Higher is better
- Comparison cohort
- artificial-analysis-automationbench-aa:AA private 657-task split of AutomationBench v1.0.6:v5.0.1:reasoning:effort=xhigh:suffix=Xhigh%2C%20Default%20Fallback
- Source type
- independent-evaluator
- Evaluator
- Artificial Analysis
- Status
- active
- Retrieved at
- 2026-11-09
- Methodology
- Exact AA configuration: Claude Fable 5.2 (Xhigh, Default Fallback) (release claude-fable-5-2, row claude-fable-5-2-xhigh, UUID 76942526-cc96-46b0-bc27-29561c53ed70, reasoning=true, effort=xhigh, suffix=Xhigh, Default Fallback). This source capture does not provide an average reasoning-token count.. Source model ID: artificial-analysis/76942526-cc96-46b0-bc27-29561c53ed70/claude-fable-5-2/claude-fable-5-2-xhigh. Source model name: Claude Fable 5.2. Provider reported by source: Anthropic. Model snapshot: Claude Fable 5.2 (Xhigh, Default Fallback). Evaluator: Artificial Analysis. Dataset type: artificial-analysis-evaluation. Reasoning: reasoning enabled. Reasoning effort: xhigh. Score display: 74.2%. Source ID: artificial-analysis. Evidence capture ID: aa-model-page-2026-11-09-7542f902e8e4. Source payload hash: 7542f902e8e4f947d83dc15004997367540a045beb71bef86f98e77adbc6a23a. Parser version: artificial-analysis-rsc-models-v4. Imported at: 2026-11-09
- Comparison methodology
- Exact AA configuration: Claude Fable 5.2 (Xhigh, Default Fallback) (release claude-fable-5-2, row claude-fable-5-2-xhigh, UUID 76942526-cc96-46b0-bc27-29561c53ed70, reasoning=true, effort=xhigh, suffix=Xhigh, Default Fallback). This source capture does not provide an average reasoning-token count.. Reasoning: reasoning enabled. Reasoning effort: xhigh
AutomationBench-AA: Overall (Show measurement, test conditions, and source)
- Source value
- 67.99749206673704
- Metric
- Objectives completed without guardrail violations
- Unit
- %
- Benchmark version
- AA private 657-task split of AutomationBench v1.0.6
- Category
- agentic
- Task
- overall
- Task label
- Overall
- Direction
- Higher is better
- Comparison cohort
- artificial-analysis-automationbench-aa:AA private 657-task split of AutomationBench v1.0.6:v5.0.1:reasoning:effort=low:suffix=Low%2C%20Default%20Fallback
- Source type
- independent-evaluator
- Evaluator
- Artificial Analysis
- Status
- active
- Retrieved at
- 2026-11-09
- Methodology
- Exact AA configuration: Claude Fable 5.2 (Low, Default Fallback) (release claude-fable-5-2, row claude-fable-5-2-low, UUID cfc7acb1-d903-4858-9b6c-6c7776caf8f9, reasoning=true, effort=low, suffix=Low, Default Fallback). This source capture does not provide an average reasoning-token count.. Source model ID: artificial-analysis/cfc7acb1-d903-4858-9b6c-6c7776caf8f9/claude-fable-5-2/claude-fable-5-2-low. Source model name: Claude Fable 5.2. Provider reported by source: Anthropic. Model snapshot: Claude Fable 5.2 (Low, Default Fallback). Evaluator: Artificial Analysis. Dataset type: artificial-analysis-evaluation. Reasoning: reasoning enabled. Reasoning effort: low. Score display: 68.0%. Source ID: artificial-analysis. Evidence capture ID: aa-model-page-2026-11-09-7542f902e8e4. Source payload hash: 7542f902e8e4f947d83dc15004997367540a045beb71bef86f98e77adbc6a23a. Parser version: artificial-analysis-rsc-models-v4. Imported at: 2026-11-09
- Comparison methodology
- Exact AA configuration: Claude Fable 5.2 (Low, Default Fallback) (release claude-fable-5-2, row claude-fable-5-2-low, UUID cfc7acb1-d903-4858-9b6c-6c7776caf8f9, reasoning=true, effort=low, suffix=Low, Default Fallback). This source capture does not provide an average reasoning-token count.. Reasoning: reasoning enabled. Reasoning effort: low
AutomationBench-AA: Overall (Show measurement, test conditions, and source)
- Source value
- 71.35975994544773
- Metric
- Objectives completed without guardrail violations
- Unit
- %
- Benchmark version
- AA private 657-task split of AutomationBench v1.0.6
- Category
- agentic
- Task
- overall
- Task label
- Overall
- Direction
- Higher is better
- Comparison cohort
- artificial-analysis-automationbench-aa:AA private 657-task split of AutomationBench v1.0.6:v5.0.1:reasoning:effort=high:suffix=High%2C%20Default%20Fallback
- Source type
- independent-evaluator
- Evaluator
- Artificial Analysis
- Status
- active
- Retrieved at
- 2026-11-09
- Methodology
- Exact AA configuration: Claude Fable 5.2 (High, Default Fallback) (release claude-fable-5-2, row claude-fable-5-2-high, UUID fdc1913b-9497-45ef-83f7-17dae756f37e, reasoning=true, effort=high, suffix=High, Default Fallback). This source capture does not provide an average reasoning-token count.. Source model ID: artificial-analysis/fdc1913b-9497-45ef-83f7-17dae756f37e/claude-fable-5-2/claude-fable-5-2-high. Source model name: Claude Fable 5.2. Provider reported by source: Anthropic. Model snapshot: Claude Fable 5.2 (High, Default Fallback). Evaluator: Artificial Analysis. Dataset type: artificial-analysis-evaluation. Reasoning: reasoning enabled. Reasoning effort: high. Score display: 71.4%. Source ID: artificial-analysis. Evidence capture ID: aa-model-page-2026-11-09-7542f902e8e4. Source payload hash: 7542f902e8e4f947d83dc15004997367540a045beb71bef86f98e77adbc6a23a. Parser version: artificial-analysis-rsc-models-v4. Imported at: 2026-11-09
- Comparison methodology
- Exact AA configuration: Claude Fable 5.2 (High, Default Fallback) (release claude-fable-5-2, row claude-fable-5-2-high, UUID fdc1913b-9497-45ef-83f7-17dae756f37e, reasoning=true, effort=high, suffix=High, Default Fallback). This source capture does not provide an average reasoning-token count.. Reasoning: reasoning enabled. Reasoning effort: high
CritPt v2 (Hard Subset): Overall (Show measurement, test conditions, and source)
- Source value
- 61.5
- Metric
- pass@1 accuracy
- Unit
- %
- Benchmark version
- v2 hard subset, November 2026 grading
- Category
- reasoning
- Task
- overall
- Task label
- Overall
- Direction
- Higher is better
- Comparison cohort
- artificial-analysis-critpt-v2-hard-subset:v2 hard subset, November 2026 grading:v5.0.1:reasoning:effort=medium:suffix=Medium%2C%20Default%20Fallback
- Source type
- independent-evaluator
- Evaluator
- Artificial Analysis
- Status
- active
- Retrieved at
- 2026-11-09
- Methodology
- Exact AA configuration: Claude Fable 5.2 (Medium, Default Fallback) (release claude-fable-5-2, row claude-fable-5-2-medium, UUID 0c9b4e05-9259-42cc-b40e-d6acbbeaa7e8, reasoning=true, effort=medium, suffix=Medium, Default Fallback). This source capture does not provide an average reasoning-token count.. Source model ID: artificial-analysis/0c9b4e05-9259-42cc-b40e-d6acbbeaa7e8/claude-fable-5-2/claude-fable-5-2-medium. Source model name: Claude Fable 5.2. Provider reported by source: Anthropic. Model snapshot: Claude Fable 5.2 (Medium, Default Fallback). Evaluator: Artificial Analysis. Dataset type: artificial-analysis-evaluation. Reasoning: reasoning enabled. Reasoning effort: medium. Score display: 61.5%. Source ID: artificial-analysis. Evidence capture ID: aa-model-page-2026-11-09-7542f902e8e4. Source payload hash: 7542f902e8e4f947d83dc15004997367540a045beb71bef86f98e77adbc6a23a. Parser version: artificial-analysis-rsc-models-v4. Imported at: 2026-11-09
- Comparison methodology
- Exact AA configuration: Claude Fable 5.2 (Medium, Default Fallback) (release claude-fable-5-2, row claude-fable-5-2-medium, UUID 0c9b4e05-9259-42cc-b40e-d6acbbeaa7e8, reasoning=true, effort=medium, suffix=Medium, Default Fallback). This source capture does not provide an average reasoning-token count.. Reasoning: reasoning enabled. Reasoning effort: medium
- Context
- This is the 40-question hard subset of CritPt v2, not the full 70-question CritPt v2 run.
CritPt v2 (Hard Subset): Overall (Show measurement, test conditions, and source)
- Source value
- 76.5
- Metric
- pass@1 accuracy
- Unit
- %
- Benchmark version
- v2 hard subset, November 2026 grading
- Category
- reasoning
- Task
- overall
- Task label
- Overall
- Direction
- Higher is better
- Comparison cohort
- artificial-analysis-critpt-v2-hard-subset:v2 hard subset, November 2026 grading:v5.0.1:reasoning:effort=max:suffix=Max%2C%20Default%20Fallback
- Source type
- independent-evaluator
- Evaluator
- Artificial Analysis
- Status
- active
- Retrieved at
- 2026-11-09
- Methodology
- Exact AA configuration: Claude Fable 5.2 (Max, Default Fallback) (release claude-fable-5-2, row claude-fable-5-2, UUID 593f2d23-247b-41c9-a679-cebd225f0a9a, reasoning=true, effort=max, suffix=Max, Default Fallback). This source capture does not provide an average reasoning-token count.. Source model ID: artificial-analysis/593f2d23-247b-41c9-a679-cebd225f0a9a/claude-fable-5-2/claude-fable-5-2. Source model name: Claude Fable 5.2. Provider reported by source: Anthropic. Model snapshot: Claude Fable 5.2 (Max, Default Fallback). Evaluator: Artificial Analysis. Dataset type: artificial-analysis-evaluation. Reasoning: reasoning enabled. Reasoning effort: max. Score display: 76.5%. Source ID: artificial-analysis. Evidence capture ID: aa-model-page-2026-11-09-7542f902e8e4. Source payload hash: 7542f902e8e4f947d83dc15004997367540a045beb71bef86f98e77adbc6a23a. Parser version: artificial-analysis-rsc-models-v4. Imported at: 2026-11-09
- Comparison methodology
- Exact AA configuration: Claude Fable 5.2 (Max, Default Fallback) (release claude-fable-5-2, row claude-fable-5-2, UUID 593f2d23-247b-41c9-a679-cebd225f0a9a, reasoning=true, effort=max, suffix=Max, Default Fallback). This source capture does not provide an average reasoning-token count.. Reasoning: reasoning enabled. Reasoning effort: max
- Context
- This is the 40-question hard subset of CritPt v2, not the full 70-question CritPt v2 run.
CritPt v2 (Hard Subset): Overall (Show measurement, test conditions, and source)
- Source value
- 73.5
- Metric
- pass@1 accuracy
- Unit
- %
- Benchmark version
- v2 hard subset, November 2026 grading
- Category
- reasoning
- Task
- overall
- Task label
- Overall
- Direction
- Higher is better
- Comparison cohort
- artificial-analysis-critpt-v2-hard-subset:v2 hard subset, November 2026 grading:v5.0.1:reasoning:effort=xhigh:suffix=Xhigh%2C%20Default%20Fallback
- Source type
- independent-evaluator
- Evaluator
- Artificial Analysis
- Status
- active
- Retrieved at
- 2026-11-09
- Methodology
- Exact AA configuration: Claude Fable 5.2 (Xhigh, Default Fallback) (release claude-fable-5-2, row claude-fable-5-2-xhigh, UUID 76942526-cc96-46b0-bc27-29561c53ed70, reasoning=true, effort=xhigh, suffix=Xhigh, Default Fallback). This source capture does not provide an average reasoning-token count.. Source model ID: artificial-analysis/76942526-cc96-46b0-bc27-29561c53ed70/claude-fable-5-2/claude-fable-5-2-xhigh. Source model name: Claude Fable 5.2. Provider reported by source: Anthropic. Model snapshot: Claude Fable 5.2 (Xhigh, Default Fallback). Evaluator: Artificial Analysis. Dataset type: artificial-analysis-evaluation. Reasoning: reasoning enabled. Reasoning effort: xhigh. Score display: 73.5%. Source ID: artificial-analysis. Evidence capture ID: aa-model-page-2026-11-09-7542f902e8e4. Source payload hash: 7542f902e8e4f947d83dc15004997367540a045beb71bef86f98e77adbc6a23a. Parser version: artificial-analysis-rsc-models-v4. Imported at: 2026-11-09
- Comparison methodology
- Exact AA configuration: Claude Fable 5.2 (Xhigh, Default Fallback) (release claude-fable-5-2, row claude-fable-5-2-xhigh, UUID 76942526-cc96-46b0-bc27-29561c53ed70, reasoning=true, effort=xhigh, suffix=Xhigh, Default Fallback). This source capture does not provide an average reasoning-token count.. Reasoning: reasoning enabled. Reasoning effort: xhigh
- Context
- This is the 40-question hard subset of CritPt v2, not the full 70-question CritPt v2 run.
CritPt v2 (Hard Subset): Overall (Show measurement, test conditions, and source)
- Source value
- 60
- Metric
- pass@1 accuracy
- Unit
- %
- Benchmark version
- v2 hard subset, November 2026 grading
- Category
- reasoning
- Task
- overall
- Task label
- Overall
- Direction
- Higher is better
- Comparison cohort
- artificial-analysis-critpt-v2-hard-subset:v2 hard subset, November 2026 grading:v5.0.1:reasoning:effort=low:suffix=Low%2C%20Default%20Fallback
- Source type
- independent-evaluator
- Evaluator
- Artificial Analysis
- Status
- active
- Retrieved at
- 2026-11-09
- Methodology
- Exact AA configuration: Claude Fable 5.2 (Low, Default Fallback) (release claude-fable-5-2, row claude-fable-5-2-low, UUID cfc7acb1-d903-4858-9b6c-6c7776caf8f9, reasoning=true, effort=low, suffix=Low, Default Fallback). This source capture does not provide an average reasoning-token count.. Source model ID: artificial-analysis/cfc7acb1-d903-4858-9b6c-6c7776caf8f9/claude-fable-5-2/claude-fable-5-2-low. Source model name: Claude Fable 5.2. Provider reported by source: Anthropic. Model snapshot: Claude Fable 5.2 (Low, Default Fallback). Evaluator: Artificial Analysis. Dataset type: artificial-analysis-evaluation. Reasoning: reasoning enabled. Reasoning effort: low. Score display: 60.0%. Source ID: artificial-analysis. Evidence capture ID: aa-model-page-2026-11-09-7542f902e8e4. Source payload hash: 7542f902e8e4f947d83dc15004997367540a045beb71bef86f98e77adbc6a23a. Parser version: artificial-analysis-rsc-models-v4. Imported at: 2026-11-09
- Comparison methodology
- Exact AA configuration: Claude Fable 5.2 (Low, Default Fallback) (release claude-fable-5-2, row claude-fable-5-2-low, UUID cfc7acb1-d903-4858-9b6c-6c7776caf8f9, reasoning=true, effort=low, suffix=Low, Default Fallback). This source capture does not provide an average reasoning-token count.. Reasoning: reasoning enabled. Reasoning effort: low
- Context
- This is the 40-question hard subset of CritPt v2, not the full 70-question CritPt v2 run.
CritPt v2 (Hard Subset): Overall (Show measurement, test conditions, and source)
- Source value
- 73
- Metric
- pass@1 accuracy
- Unit
- %
- Benchmark version
- v2 hard subset, November 2026 grading
- Category
- reasoning
- Task
- overall
- Task label
- Overall
- Direction
- Higher is better
- Comparison cohort
- artificial-analysis-critpt-v2-hard-subset:v2 hard subset, November 2026 grading:v5.0.1:reasoning:effort=high:suffix=High%2C%20Default%20Fallback
- Source type
- independent-evaluator
- Evaluator
- Artificial Analysis
- Status
- active
- Retrieved at
- 2026-11-09
- Methodology
- Exact AA configuration: Claude Fable 5.2 (High, Default Fallback) (release claude-fable-5-2, row claude-fable-5-2-high, UUID fdc1913b-9497-45ef-83f7-17dae756f37e, reasoning=true, effort=high, suffix=High, Default Fallback). This source capture does not provide an average reasoning-token count.. Source model ID: artificial-analysis/fdc1913b-9497-45ef-83f7-17dae756f37e/claude-fable-5-2/claude-fable-5-2-high. Source model name: Claude Fable 5.2. Provider reported by source: Anthropic. Model snapshot: Claude Fable 5.2 (High, Default Fallback). Evaluator: Artificial Analysis. Dataset type: artificial-analysis-evaluation. Reasoning: reasoning enabled. Reasoning effort: high. Score display: 73.0%. Source ID: artificial-analysis. Evidence capture ID: aa-model-page-2026-11-09-7542f902e8e4. Source payload hash: 7542f902e8e4f947d83dc15004997367540a045beb71bef86f98e77adbc6a23a. Parser version: artificial-analysis-rsc-models-v4. Imported at: 2026-11-09
- Comparison methodology
- Exact AA configuration: Claude Fable 5.2 (High, Default Fallback) (release claude-fable-5-2, row claude-fable-5-2-high, UUID fdc1913b-9497-45ef-83f7-17dae756f37e, reasoning=true, effort=high, suffix=High, Default Fallback). This source capture does not provide an average reasoning-token count.. Reasoning: reasoning enabled. Reasoning effort: high
- Context
- This is the 40-question hard subset of CritPt v2, not the full 70-question CritPt v2 run.
FrontierCode: Overall (Show measurement, test conditions, and source)
- Source value
- 41.2984090909091
- Metric
- Mean score over three attempts per task
- Unit
- %
- Benchmark version
- 2.0
- Category
- coding
- Task
- overall
- Task label
- Overall
- Direction
- Higher is better
- Comparison cohort
- artificial-analysis-frontiercode-2-0:2.0:v5.0.1:reasoning:effort=medium:suffix=Medium%2C%20Default%20Fallback
- Source type
- independent-evaluator
- Evaluator
- Artificial Analysis
- Status
- active
- Retrieved at
- 2026-11-09
- Methodology
- Exact AA configuration: Claude Fable 5.2 (Medium, Default Fallback) (release claude-fable-5-2, row claude-fable-5-2-medium, UUID 0c9b4e05-9259-42cc-b40e-d6acbbeaa7e8, reasoning=true, effort=medium, suffix=Medium, Default Fallback). This source capture does not provide an average reasoning-token count.. Source model ID: artificial-analysis/0c9b4e05-9259-42cc-b40e-d6acbbeaa7e8/claude-fable-5-2/claude-fable-5-2-medium. Source model name: Claude Fable 5.2. Provider reported by source: Anthropic. Model snapshot: Claude Fable 5.2 (Medium, Default Fallback). Evaluator: Artificial Analysis. Dataset type: artificial-analysis-evaluation. Reasoning: reasoning enabled. Reasoning effort: medium. Score display: 41.3%. Source ID: artificial-analysis. Evidence capture ID: aa-model-page-2026-11-09-7542f902e8e4. Source payload hash: 7542f902e8e4f947d83dc15004997367540a045beb71bef86f98e77adbc6a23a. Parser version: artificial-analysis-rsc-models-v4. Imported at: 2026-11-09
- Comparison methodology
- Exact AA configuration: Claude Fable 5.2 (Medium, Default Fallback) (release claude-fable-5-2, row claude-fable-5-2-medium, UUID 0c9b4e05-9259-42cc-b40e-d6acbbeaa7e8, reasoning=true, effort=medium, suffix=Medium, Default Fallback). This source capture does not provide an average reasoning-token count.. Reasoning: reasoning enabled. Reasoning effort: medium
FrontierCode: Overall (Show measurement, test conditions, and source)
- Source value
- 48.7597222222222
- Metric
- Mean score over three attempts per task
- Unit
- %
- Benchmark version
- 2.0
- Category
- coding
- Task
- overall
- Task label
- Overall
- Direction
- Higher is better
- Comparison cohort
- artificial-analysis-frontiercode-2-0:2.0:v5.0.1:reasoning:effort=max:suffix=Max%2C%20Default%20Fallback
- Source type
- independent-evaluator
- Evaluator
- Artificial Analysis
- Status
- active
- Retrieved at
- 2026-11-09
- Methodology
- Exact AA configuration: Claude Fable 5.2 (Max, Default Fallback) (release claude-fable-5-2, row claude-fable-5-2, UUID 593f2d23-247b-41c9-a679-cebd225f0a9a, reasoning=true, effort=max, suffix=Max, Default Fallback). This source capture does not provide an average reasoning-token count.. Source model ID: artificial-analysis/593f2d23-247b-41c9-a679-cebd225f0a9a/claude-fable-5-2/claude-fable-5-2. Source model name: Claude Fable 5.2. Provider reported by source: Anthropic. Model snapshot: Claude Fable 5.2 (Max, Default Fallback). Evaluator: Artificial Analysis. Dataset type: artificial-analysis-evaluation. Reasoning: reasoning enabled. Reasoning effort: max. Score display: 48.8%. Source ID: artificial-analysis. Evidence capture ID: aa-model-page-2026-11-09-7542f902e8e4. Source payload hash: 7542f902e8e4f947d83dc15004997367540a045beb71bef86f98e77adbc6a23a. Parser version: artificial-analysis-rsc-models-v4. Imported at: 2026-11-09
- Comparison methodology
- Exact AA configuration: Claude Fable 5.2 (Max, Default Fallback) (release claude-fable-5-2, row claude-fable-5-2, UUID 593f2d23-247b-41c9-a679-cebd225f0a9a, reasoning=true, effort=max, suffix=Max, Default Fallback). This source capture does not provide an average reasoning-token count.. Reasoning: reasoning enabled. Reasoning effort: max
Page 1 of 25
Claude Fable 5.2 compared
Predecessor and successor
The most popular comparisons
- 1Claude Fable 5.2 vs. Kimi K3
- 2Claude Fable 5.2 vs. Claude Opus 5.5
- 3Claude Fable 5.2 vs. Claude Haiku 5.5
- 4Claude Fable 5.2 vs. Claude Sonnet 5.5
- 5Claude Fable 5.2 vs. Claude Opus 5
- 6Claude Fable 5.2 vs. GPT-6.1 Sol
- 7Claude Fable 5.2 vs. GPT-6.1 Terra
- 8Claude Fable 5.2 vs. Gemini 4 Argon
- 9Claude Fable 5.2 vs. Muse Wonder
Show all 17 comparisons
Similar models
More models of the same kind as Claude Fable 5.2, chosen by model line, developer, size, price and release date.
- Claude Fable 5.1Anthropic · same model line
- Claude Opus 5Anthropic · same developer
- Claude Opus 5.5Anthropic · same developer
- Claude Fable 5Anthropic · same developer
- Claude Mythos 5Anthropic · same developer
- Claude Mythos 5.1Anthropic · same developer
- Claude Opus 4.7Anthropic · same developer
- Claude Opus 4.6Anthropic · same developer
More about Claude Fable 5.2
All specifications
Architecture
Input and output
- Structured output
- JSON outputs and strict tool use are supported. Structured outputs are unavailable for organizations using customer-managed encryption keys.
- Source: Structured output
Watermarking and provenance
AI watermarking refers to invisible signals embedded in the output. Visible marks and provenance metadata are listed separately. Unknown means there is insufficient evidence for that particular field.
- AI watermarking
- Unknown
- no source
- Watermarking technique
- Not known
- no source
Licence and access
- Shutdown
- None announced
- no source
API
Further details
- Notes
- For demanding research, coding, and long-running agent tasks. Adaptive thinking is always on, with high effort as the API default. Parameter count and architecture are undisclosed.
- Source: Notes
- Reasoning
- Adaptive thinking is always on and cannot be disabled. Effort levels are low, medium, high, xhigh, and max. The API default is high.
- Source: Reasoning
- Sampling parameters
- temperature, top_p, and top_k are unsupported. Forced tool use through tool_choice any or tool is also unsupported.
- Source: Sampling parameters
- Other API access
- Amazon Bedrock uses anthropic.claude-fable-5-2. Claude Platform on AWS, Claude Models on Google Cloud, and Microsoft Foundry document claude-fable-5-2.
- Source: Other API access
Frequently asked questions about Claude Fable 5.2
How good is Claude Fable 5.2 in benchmarks?
Very good: Rank 1 of 94 in the overall index. First place is not certain (win chance 73%): Gemini 4 Carbon and Claude Opus 5.5 are statistically level.
Overall index (Gradually Meta-Score v5) from 26 benchmark results · data through Nov 10, 2026
Claude Fable 5.2Rank 1130.9
Gemini 4 Carbon127.0
Claude Opus 5.5125.6
Claude Fable 5.1124.1
Claude Sonnet 5.5121.8
Gemini 4 Argon121.4
GPT-6 Astra120.2
What is the Overall index (Gradually Meta-Score v5)?
A point value that combines many independent benchmarks into one number. 100 points is the average of a fixed group of 29 models, and a model 15 points ahead does better in roughly three of four benchmark comparisons. Claude Fable 5.2 scores 130.9 points, which is rank 1 of 94.
Own analysis by gradually.ai, method benchmark-meta-score-v5 · for Claude Fable 5.2 from 26 test series
How much does Claude Fable 5.2 cost?
Through the API $8 per 1M input tokens and $40 output. That is expensive.
Prices from Nov 11, 2026
Nov 10, 2026Nov 11, 2026
Unchanged since tracking began on Nov 10, 2026.
What is the knowledge cutoff of Claude Fable 5.2?
June 2026, according to Anthropic. That is 5 months before release.
Stated by Anthropic
JunJulAugSepOctNovKnowledge cutoffJune 2026ReleasedNov 9, 2026Claude Fable 5.1 cutoffJune 2026
How large is the context window of Claude Fable 5.2?
1 M tokens. That is how much text Claude Fable 5.2 reads at once, about 1,500 A4 pages.
Page count as a rule of thumb from Artificial Analysis
GPT-6 Astra1.05M
Gemini 4 Argon1.05M
Claude Fable 5.2Rank 36 of 1101M
Gemini 4 Carbon1M
Claude Opus 5.51M
Claude Fable 5.11M
Claude Sonnet 5.51M
Can I run Claude Fable 5.2 locally?
No. Anthropic does not publish the weights.
Stated by Anthropic
- Open weights
- no
- License
- n/a
- Parameters
- n/a
- active per token
- n/a
Where can I use Claude Fable 5.2?
At Anthropic with a subscription from $20 a month. Including other providers that makes 10 plans in Germany, plus the API.
Subscription comparison, prices in Germany incl. VAT
At Anthropic
5 plans, from $20 a month (Claude Pro).
Through other providers
5 plans from 2 providers, from $39 a month (GitHub Copilot Pro+), with limited credits.
Through the API
$8 input, $40 output per 1M tokens.
Is Claude Fable 5.2 better than Claude Fable 5.1?
Yes, clearly in the overall index. It leads in 6 of 8 areas tested for both.
Claude Fable 5.1 as predecessor according to Anthropic: Claude Fable 5.2 migration guide, checked Nov 9, 2026
130.9
Claude Fable 5.2
Rank 1
124.1
Claude Fable 5.1
Rank 4
n/a
Claude Fable 5.2
59
Claude Fable 5.1
fairly slow
$16
Claude Fable 5.2
cheaper
$20
Claude Fable 5.1
1.3 times the price
AreaFable 5.2Fable 5.1100 points = average, higher is betterResult
CodingFable 5.2 aheadFable 5.2: 131.6Fable 5.1: 122.8
ReasoningFable 5.2 aheadFable 5.2: 124.9Fable 5.1: 116.1
MathematicsFable 5.2 aheadFable 5.2: 135.2Fable 5.1: 132.5
KnowledgeFable 5.2 aheadFable 5.2: 145.5Fable 5.1: 130.1
AgentsFable 5.2 aheadFable 5.2: 124.2Fable 5.1: 114.0
GeneralFable 5.2 aheadFable 5.2: 132.8Fable 5.1: 122.8
Financelevel3 and 6 testsFable 5.2: 123.5Fable 5.1: 121.8
LegalFable 5.1 aheadFable 5.2: 120.9Fable 5.1: 130.6
Healthonly Fable 5.1 has a scoreFable 5.2: not enough dataFable 5.1: 123.4
Multimodalonly Fable 5.1 has a scoreFable 5.2: not enough dataFable 5.1: 121.1
Sources and data date
Show all 38 sources
Benchmark sources at a glance
Which independent evaluators tested this model. This is coverage, not a quality ranking.
| Evaluator | Test series | Composite indices | Recorded active measurements |
|---|---|---|---|
| Artificial Analysis | 11 | 0 | 65 |
| Vals AI | 15 | 4 | 176 |
| LiveBench | 7 | 1 | 41 |
Without vendor claims and historical or disputed results; those stay labelled in the single values. 1 measurements are not yet assigned to a test series.
| Type | Evidence and data date |
|---|---|
| Price | Base API prices without caching or batch discounts. Higher context tiers and other rates are listed under Costs. |
| Input | Documented modalities. Retrieved November 9, 2026. |
| Output | Documented modalities. Retrieved November 9, 2026. |
| Research status | Research date November 10, 2026. 7 source URLs checked. This documents the inspected sources, not an exhaustive inventory of every publication. |
| Additional source | Release announcement (retrieved November 9, 2026) · Anthropic: Claude Fable and Mythos 5.2 · Editorial description reviewed November 9, 2026 |
| Additional source | Anthropic: What’s new in Claude Fable 5.2 (retrieved November 9, 2026) · What’s new in Claude Fable 5.2 · Editorial description reviewed November 9, 2026 |
| Additional source | Artificial Analysis (retrieved November 9, 2026; November 10, 2026)12 |
| Additional source | Artificial Analysis (retrieved November 9, 2026)12345678910 |
| Additional source | Vals AI (retrieved November 10, 2026)12345678910111213141516171819 |
| Additional source | LiveBench (retrieved November 10, 2026)12 |
| Additional source | Model metadata (retrieved November 10, 2026) |
| Additional source | Anthropic: Claude Fable 5.2 migration guide (2026-11-09) |