xAILanguage modelActive
Grok 4.20 Non-Reasoning
Grok 4.20 Non-Reasoning has no overall index value yet because there are too few independent benchmark results.
Intelligencenot ranked yet
n/aOverall indexhigher is better
SpeedRank 48 of 101
88Tokens per secondhigher is better
Priceno API price
n/aper 1M tokenslower is better
ContextRank 31 of 102
1Mtokens at once≈ 1,500 A4 pages
PopularityRank 670 of 875
0out of 100same as last month
Summary
At 88 tokens per second it writes faster than a typical model (87).
Grok 4.20 Non-Reasoning has its own API identifier, grok-4.20-0309-non-reasoning, alongside Grok 4.20 reasoning. xAI lists them as separate variants. The Non-Reasoning version accepts text and images with up to 1 million context tokens and produces text.
All benchmark results
Every published result with its source, test setup and comparison models.
Show all 15 published benchmark results
Page 1 of 2
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.
CritPt v2: Overall (Show measurement, test conditions, and source)
- Source value
- 0.28571428571428603
- Score
- 0.28571428571428603
- Metric
- pass@1 accuracy
- Unit
- %
- Benchmark version
- v2
- Category
- reasoning
- Task
- overall
- Task label
- Overall
- Direction
- Higher is better
- Source type
- independent-evaluator
- Evaluator
- Artificial Analysis
- Status
- disputed
- Retrieved at
- 2026-11-02
- Methodology
- Exact AA configuration: Grok 4.20 0309 v2 (Non-reasoning) (release grok-4-20, row grok-4-20-non-reasoning, UUID 64312f37-3701-4243-a4c9-7c07a58cd6b9, reasoning=false, effort=not specified, suffix=Non-reasoning). AA marks this model variant as non-reasoning and reports 0 reasoning tokens. The full CritPt v2 run covers 70 questions with five repeats. The official grading server reports pass@1.. Source model ID: artificial-analysis/64312f37-3701-4243-a4c9-7c07a58cd6b9/grok-4-20/grok-4-20-non-reasoning. Source model name: Grok 4.20 0309 v2. Provider reported by source: SpaceXAI. Model snapshot: Grok 4.20 0309 v2 (Non-reasoning). Evaluator: Artificial Analysis. Dataset type: artificial-analysis-evaluation. Tool policy: Python functions, symbolic expressions, and numerical answers; official grading server. Attempts: 70 questions, five repeats per question, pass@1. Reasoning: non-reasoning. Reasoning effort: none. Score display: 0.3%. Source ID: artificial-analysis. Evidence capture ID: aa-model-page-2026-11-02-67d46be20a5b. Source payload hash: 67d46be20a5b412db6cfb60ad190441fcaa5513500b8fd47d033168d0e97b6b2. Parser version: artificial-analysis-rsc-models-v3. Imported at: 2026-11-02. Source cohort: artificial-analysis-critpt-v2:v2:standalone:non-reasoning:effort=unspecified:suffix=Non-reasoning
- Comparison methodology
- Exact AA configuration: Grok 4.20 0309 v2 (Non-reasoning) (release grok-4-20, row grok-4-20-non-reasoning, UUID 64312f37-3701-4243-a4c9-7c07a58cd6b9, reasoning=false, effort=not specified, suffix=Non-reasoning). AA marks this model variant as non-reasoning and reports 0 reasoning tokens. The full CritPt v2 run covers 70 questions with five repeats. The official grading server reports pass@1.. Tool policy: Python functions, symbolic expressions, and numerical answers; official grading server. Attempts: 70 questions, five repeats per question, pass@1. Reasoning: non-reasoning. Reasoning effort: none
- Context
- This AA source field contains the original CritPt score. The earlier CritPt v2 attribution was incorrect; the value and former label remain as disputed historical evidence.. The source dataset marks this result as disputed. It remains visible as documented evidence, but is not reliable comparative proof.
IFBench: Overall (Show measurement, test conditions, and source)
- Source value
- 49.319727891156504
- Score
- 49.319727891156504
- Metric
- IFBench prompt-level loose accuracy (pass@1)
- Unit
- %
- Benchmark version
- AA test set: 294 questions
- Category
- general
- Task
- overall
- Task label
- Overall
- Direction
- Higher is better
- Source type
- independent-evaluator
- Evaluator
- Artificial Analysis
- Status
- superseded
- Retrieved at
- 2026-11-02
- Methodology
- Exact AA configuration: Grok 4.20 0309 v2 (Non-reasoning) (release grok-4-20, row grok-4-20-non-reasoning, UUID 64312f37-3701-4243-a4c9-7c07a58cd6b9, reasoning=false, effort=not specified, suffix=Non-reasoning). AA marks this model variant as non-reasoning and reports 0 reasoning tokens. Artificial Analysis runs 294 questions from the IFBench test set with five repeats per question. The official evaluator scores prompt-level accuracy in loose mode as pass@1.. Source model ID: artificial-analysis/64312f37-3701-4243-a4c9-7c07a58cd6b9/grok-4-20/grok-4-20-non-reasoning. Source model name: Grok 4.20 0309 v2. Provider reported by source: SpaceXAI. Model snapshot: Grok 4.20 0309 v2 (Non-reasoning). Evaluator: Artificial Analysis. Dataset type: artificial-analysis-evaluation. Tool policy: Single-turn answer evaluated with the official IFBench source code. Attempts: pass@1, 5 repeats per question. Reasoning: non-reasoning. Reasoning effort: none. Score display: 49.3%. Source ID: artificial-analysis. Evidence capture ID: aa-model-page-2026-11-02-67d46be20a5b. Source payload hash: 67d46be20a5b412db6cfb60ad190441fcaa5513500b8fd47d033168d0e97b6b2. Parser version: artificial-analysis-rsc-models-v3. Imported at: 2026-11-02. Source cohort: artificial-analysis-ifbench:AA test set: 294 questions:standalone:non-reasoning:effort=unspecified:suffix=Non-reasoning
- Comparison methodology
- Exact AA configuration: Grok 4.20 0309 v2 (Non-reasoning) (release grok-4-20, row grok-4-20-non-reasoning, UUID 64312f37-3701-4243-a4c9-7c07a58cd6b9, reasoning=false, effort=not specified, suffix=Non-reasoning). AA marks this model variant as non-reasoning and reports 0 reasoning tokens. Artificial Analysis runs 294 questions from the IFBench test set with five repeats per question. The official evaluator scores prompt-level accuracy in loose mode as pass@1.. Tool policy: Single-turn answer evaluated with the official IFBench source code. Attempts: pass@1, 5 repeats per question. Reasoning: non-reasoning. Reasoning effort: none
- Context
- The often cited 58 refers to verifiable out-of-domain constraint types, not the number of questions in AA's test run.. This result is retained as historical evidence and is not interchangeable with current results.
IFBench: Overall (Show measurement, test conditions, and source)
- Source value
- 49.319727891156504
- Score
- 49.319727891156504
- Metric
- IFBench prompt-level loose accuracy (pass@1)
- Unit
- %
- Benchmark version
- AA test set: 294 questions
- Category
- general
- Task
- overall
- Task label
- Overall
- Direction
- Higher is better
- Source type
- independent-evaluator
- Evaluator
- Artificial Analysis
- Status
- active
- Retrieved at
- 2026-11-06
- Methodology
- Exact AA configuration: Grok 4.20 0309 v2 (Non-reasoning) (release grok-4-20, row grok-4-20-non-reasoning, UUID 64312f37-3701-4243-a4c9-7c07a58cd6b9, reasoning=false, effort=not specified, suffix=Non-reasoning). This source capture does not provide an average reasoning-token count. Artificial Analysis runs 294 questions from the IFBench test set with five repeats per question. The official evaluator scores prompt-level accuracy in loose mode as pass@1.. Source model ID: artificial-analysis/64312f37-3701-4243-a4c9-7c07a58cd6b9/grok-4-20/grok-4-20-non-reasoning. Source model name: Grok 4.20 0309 v2. Provider reported by source: SpaceXAI. Model snapshot: Grok 4.20 0309 v2 (Non-reasoning). Evaluator: Artificial Analysis. Dataset type: artificial-analysis-evaluation. Tool policy: Single-turn answer evaluated with the official IFBench source code. Attempts: pass@1, 5 repeats per question. Reasoning: non-reasoning. Reasoning effort: none. Score display: 49.3%. Source ID: artificial-analysis. Evidence capture ID: aa-model-page-2026-11-06-ec1194e008b3. Source payload hash: ec1194e008b3638ef60dbf5249daf9e6047364e22905e745a4af64c5f3f080fe. Parser version: artificial-analysis-rsc-models-v4. Imported at: 2026-11-06. Source cohort: artificial-analysis-ifbench:AA test set: 294 questions:standalone:non-reasoning:effort=unspecified:suffix=Non-reasoning
- Comparison methodology
- Exact AA configuration: Grok 4.20 0309 v2 (Non-reasoning) (release grok-4-20, row grok-4-20-non-reasoning, UUID 64312f37-3701-4243-a4c9-7c07a58cd6b9, reasoning=false, effort=not specified, suffix=Non-reasoning). This source capture does not provide an average reasoning-token count. Artificial Analysis runs 294 questions from the IFBench test set with five repeats per question. The official evaluator scores prompt-level accuracy in loose mode as pass@1.. Tool policy: Single-turn answer evaluated with the official IFBench source code. Attempts: pass@1, 5 repeats per question. Reasoning: non-reasoning. Reasoning effort: none
- Context
- The often cited 58 refers to verifiable out-of-domain constraint types, not the number of questions in AA's test run.
MMMU-Pro: Overall (Show measurement, test conditions, and source)
- Source value
- 64.91329479768791
- Score
- 64.91329479768791
- Metric
- pass@1 accuracy
- Unit
- %
- Benchmark version
- AA cohort: 1,730 questions
- Category
- multimodal
- Task
- overall
- Task label
- Overall
- Direction
- Higher is better
- Source type
- independent-evaluator
- Evaluator
- Artificial Analysis
- Status
- superseded
- Retrieved at
- 2026-11-02
- Methodology
- Exact AA configuration: Grok 4.20 0309 v2 (Non-reasoning) (release grok-4-20, row grok-4-20-non-reasoning, UUID 64312f37-3701-4243-a4c9-7c07a58cd6b9, reasoning=false, effort=not specified, suffix=Non-reasoning). AA marks this model variant as non-reasoning and reports 0 reasoning tokens. The AA run covers 1,730 questions with ten answer options. Answers are extracted by regex and scored pass@1.. Source model ID: artificial-analysis/64312f37-3701-4243-a4c9-7c07a58cd6b9/grok-4-20/grok-4-20-non-reasoning. Source model name: Grok 4.20 0309 v2. Provider reported by source: SpaceXAI. Model snapshot: Grok 4.20 0309 v2 (Non-reasoning). Evaluator: Artificial Analysis. Dataset type: artificial-analysis-evaluation. Tool policy: Ten-option multiple choice; regex answer extraction. Attempts: pass@1, one run per question. Reasoning: non-reasoning. Reasoning effort: none. Score display: 64.9%. Source ID: artificial-analysis. Evidence capture ID: aa-model-page-2026-11-02-67d46be20a5b. Source payload hash: 67d46be20a5b412db6cfb60ad190441fcaa5513500b8fd47d033168d0e97b6b2. Parser version: artificial-analysis-rsc-models-v3. Imported at: 2026-11-02. Source cohort: artificial-analysis-mmmu-pro:AA cohort: 1,730 questions:standalone:non-reasoning:effort=unspecified:suffix=Non-reasoning
- Comparison methodology
- Exact AA configuration: Grok 4.20 0309 v2 (Non-reasoning) (release grok-4-20, row grok-4-20-non-reasoning, UUID 64312f37-3701-4243-a4c9-7c07a58cd6b9, reasoning=false, effort=not specified, suffix=Non-reasoning). AA marks this model variant as non-reasoning and reports 0 reasoning tokens. The AA run covers 1,730 questions with ten answer options. Answers are extracted by regex and scored pass@1.. Tool policy: Ten-option multiple choice; regex answer extraction. Attempts: pass@1, one run per question. Reasoning: non-reasoning. Reasoning effort: none
- Context
- The benchmark page describes 3,460 questions in the full MMMU-Pro dataset; AA's methodology specifies a 1,730-question run.. This result is retained as historical evidence and is not interchangeable with current results.
MMMU-Pro: Overall (Show measurement, test conditions, and source)
- Source value
- 64.91329479768791
- Score
- 64.91329479768791
- Metric
- pass@1 accuracy
- Unit
- %
- Benchmark version
- AA cohort: 1,730 questions
- Category
- multimodal
- Task
- overall
- Task label
- Overall
- Direction
- Higher is better
- Source type
- independent-evaluator
- Evaluator
- Artificial Analysis
- Status
- active
- Retrieved at
- 2026-11-06
- Methodology
- Exact AA configuration: Grok 4.20 0309 v2 (Non-reasoning) (release grok-4-20, row grok-4-20-non-reasoning, UUID 64312f37-3701-4243-a4c9-7c07a58cd6b9, reasoning=false, effort=not specified, suffix=Non-reasoning). This source capture does not provide an average reasoning-token count. The AA run covers 1,730 questions with ten answer options. Answers are extracted by regex and scored pass@1.. Source model ID: artificial-analysis/64312f37-3701-4243-a4c9-7c07a58cd6b9/grok-4-20/grok-4-20-non-reasoning. Source model name: Grok 4.20 0309 v2. Provider reported by source: SpaceXAI. Model snapshot: Grok 4.20 0309 v2 (Non-reasoning). Evaluator: Artificial Analysis. Dataset type: artificial-analysis-evaluation. Tool policy: Ten-option multiple choice; regex answer extraction. Attempts: pass@1, one run per question. Reasoning: non-reasoning. Reasoning effort: none. Score display: 64.9%. Source ID: artificial-analysis. Evidence capture ID: aa-model-page-2026-11-06-ec1194e008b3. Source payload hash: ec1194e008b3638ef60dbf5249daf9e6047364e22905e745a4af64c5f3f080fe. Parser version: artificial-analysis-rsc-models-v4. Imported at: 2026-11-06. Source cohort: artificial-analysis-mmmu-pro:AA cohort: 1,730 questions:standalone:non-reasoning:effort=unspecified:suffix=Non-reasoning
- Comparison methodology
- Exact AA configuration: Grok 4.20 0309 v2 (Non-reasoning) (release grok-4-20, row grok-4-20-non-reasoning, UUID 64312f37-3701-4243-a4c9-7c07a58cd6b9, reasoning=false, effort=not specified, suffix=Non-reasoning). This source capture does not provide an average reasoning-token count. The AA run covers 1,730 questions with ten answer options. Answers are extracted by regex and scored pass@1.. Tool policy: Ten-option multiple choice; regex answer extraction. Attempts: pass@1, one run per question. Reasoning: non-reasoning. Reasoning effort: none
- Context
- The benchmark page describes 3,460 questions in the full MMMU-Pro dataset; AA's methodology specifies a 1,730-question run.
CDDM crop accuracy (Show measurement, test conditions, and source)
- Source value
- 72.8
- Score
- 72.8
- Metric
- crop-label exact-match accuracy
- Unit
- %
- Benchmark version
- VIDA CDDM Round-1; paper published 2026-08-11
- Category
- multimodal
- Direction
- Higher is better
- Sample
- 1219
- Source type
- independent-evaluator
- Evaluator
- Frontiers in Plant Science, VIDA+PANDA Tables 1 and 2
- Status
- active
- Retrieved at
- 2026-10-16
- Methodology
- Independent evaluation. Exact paper model string grok-4.20-non-reasoning, labelled Grok-4.20-Fast in Table 1. VIDA Round-1 independently evaluated 1,219 CDDM leaf images across 53 classes in a zero-shot setup. The prompt supplied the crop and disease taxonomy but withheld the dataset name; output budget was approximately 600-700 tokens. This is a per-model result, not the later PANDA debate or ensemble. Evaluation request date, temperature, and immutable serving snapshot are not reported.
- Context
- This is a task-specific zero-shot crop-disease VLM observation, not a general model-quality score. The study reports an alias string, not an immutable model snapshot or request timestamp. xAI currently documents this alias as grok-4.20-0309-non-reasoning, but the paper does not establish which snapshot served its earlier evaluation.
CDDM combined crop-disease accuracy (Show measurement, test conditions, and source)
- Source value
- 32.6
- Score
- 32.6
- Metric
- crop and disease labels both correct
- Unit
- %
- Benchmark version
- VIDA CDDM Round-1; paper published 2026-08-11
- Category
- multimodal
- Direction
- Higher is better
- Sample
- 1219
- Confidence interval
- 2.6
- Lower confidence bound
- 30
- Upper confidence bound
- 35.2
- Source type
- independent-evaluator
- Evaluator
- Frontiers in Plant Science, VIDA+PANDA Tables 1 and 2
- Status
- active
- Retrieved at
- 2026-10-16
- Methodology
- Independent evaluation. Exact paper model string grok-4.20-non-reasoning, labelled Grok-4.20-Fast in Table 1. VIDA Round-1 independently evaluated 1,219 CDDM leaf images across 53 classes in a zero-shot setup. The prompt supplied the crop and disease taxonomy but withheld the dataset name; output budget was approximately 600-700 tokens. This is a per-model result, not the later PANDA debate or ensemble. Evaluation request date, temperature, and immutable serving snapshot are not reported.
- Context
- This is a task-specific zero-shot crop-disease VLM observation, not a general model-quality score. The study reports an alias string, not an immutable model snapshot or request timestamp. xAI currently documents this alias as grok-4.20-0309-non-reasoning, but the paper does not establish which snapshot served its earlier evaluation.
CDDM disease accuracy (Show measurement, test conditions, and source)
- Source value
- 39.9
- Score
- 39.9
- Metric
- disease-label exact-match accuracy
- Unit
- %
- Benchmark version
- VIDA CDDM Round-1; paper published 2026-08-11
- Category
- multimodal
- Direction
- Higher is better
- Sample
- 1219
- Source type
- independent-evaluator
- Evaluator
- Frontiers in Plant Science, VIDA+PANDA Tables 1 and 2
- Status
- active
- Retrieved at
- 2026-10-16
- Methodology
- Independent evaluation. Exact paper model string grok-4.20-non-reasoning, labelled Grok-4.20-Fast in Table 1. VIDA Round-1 independently evaluated 1,219 CDDM leaf images across 53 classes in a zero-shot setup. The prompt supplied the crop and disease taxonomy but withheld the dataset name; output budget was approximately 600-700 tokens. This is a per-model result, not the later PANDA debate or ensemble. Evaluation request date, temperature, and immutable serving snapshot are not reported.
- Context
- This is a task-specific zero-shot crop-disease VLM observation, not a general model-quality score. The study reports an alias string, not an immutable model snapshot or request timestamp. xAI currently documents this alias as grok-4.20-0309-non-reasoning, but the paper does not establish which snapshot served its earlier evaluation.
CDDM parse failure rate (Show measurement, test conditions, and source)
- Source value
- 0
- Score
- 0
- Metric
- three-stage output parser failure rate; lower is better
- Unit
- %
- Benchmark version
- VIDA CDDM Round-1; paper published 2026-08-11
- Category
- multimodal
- Direction
- Lower is better
- Sample
- 1219
- Source type
- independent-evaluator
- Evaluator
- Frontiers in Plant Science, VIDA+PANDA Tables 1 and 2
- Status
- active
- Retrieved at
- 2026-10-16
- Methodology
- Independent evaluation. Exact paper model string grok-4.20-non-reasoning, labelled Grok-4.20-Fast in Table 1. VIDA Round-1 independently evaluated 1,219 CDDM leaf images across 53 classes in a zero-shot setup. Table 2 reports the share of outputs that failed the paper's three-stage response parser. The prompt supplied the crop and disease taxonomy but withheld the dataset name; output budget was approximately 600-700 tokens. Evaluation request date, temperature, and immutable serving snapshot are not reported.
- Context
- This records parser compatibility in one task pipeline, not a general model-quality score. A zero failure count here does not establish general structured-output reliability. The study reports an alias string, not an immutable model snapshot or request timestamp. xAI currently documents this alias as grok-4.20-0309-non-reasoning, but the paper does not establish which snapshot served its earlier evaluation.
PANDA Round 3 per-agent combined accuracy (Show measurement, test conditions, and source)
- Source value
- 32.2
- Score
- 32.2
- Metric
- Share of images where crop and disease labels are both correct after debate round 3
- Unit
- %
- Benchmark version
- VIDA CDDM evaluation, PANDA headline panel, Table 7; paper published 2026-08-11
- Category
- multimodal
- Direction
- Higher is better
- Sample
- 1219
- Source type
- community-reported
- Evaluator
- A multi-agent vision-language debate framework for zero-shot crop disease diagnosis
- Status
- active
- Retrieved at
- 2026-10-18
- Methodology
- Community-reported result. PANDA headline panel evaluated on the same 1,219-image CDDM set; Grok-4.20-Fast is identified in Table 1 by the exact model string grok-4.20-non-reasoning and debates alongside GPT-4.1 and Qwen-VL. Table 7 reports the per-agent combined crop-and-disease accuracy after round 3; its value 0.322 is expressed here as 32.2%. This is the post-debate condition, not a standalone Round-1 call.
- Context
- This score is conditional on one three-agent debate panel and is not a new dataset or a standalone model evaluation. The paper reports the R1-to-R3 Wilcoxon p-value as 0.248, but no confidence interval for the R3 cell. Evaluator affiliation with the model provider was not established in the inspected source, so the observation is conservatively classified as community-reported.
Hermes Grok Bench v0.1 compatibility micro-suite (Show measurement, test conditions, and source)
- Source value
- 2
- Score
- 2
- Metric
- tasks passed
- Unit
- count
- Benchmark version
- v0.1, run 2026-05-04T081941Z
- Category
- agentic
- Direction
- Higher is better
- Sample
- 3
- Source type
- independent-evaluator
- Evaluator
- Julien Talbot Hermes Grok Bench
- Status
- active
- Retrieved at
- 2026-10-15
- Methodology
- Independent evaluation. Exact API ID grok-4.20-0309-non-reasoning; run 2026-05-04T081941Z. Single arithmetic, weather-tool and prime-code tasks; two passed, zero reasoning tokens recorded.
- Context
- Three-task historical compatibility micro-suite, not a general capability ranking. One request per task, no uncertainty interval; unrelated to our Community tests.
CDDM GPT-5-judged reasoning quality (Show measurement, test conditions, and source)
- Source value
- 7.29
- Score
- 7.29
- Metric
- mean on a 1-10 GPT-5 judge scale
- Unit
- rating
- Benchmark version
- VIDA CDDM Round-1; paper published 2026-08-11
- Category
- multimodal
- Direction
- Higher is better
- Sample
- 1219
- Source type
- independent-evaluator
- Evaluator
- Frontiers in Plant Science, VIDA+PANDA Tables 1 and 2
- Status
- active
- Retrieved at
- 2026-10-16
- Methodology
- Independent evaluation. Exact paper model string grok-4.20-non-reasoning, labelled Grok-4.20-Fast in Table 1. VIDA Round-1 independently evaluated 1,219 CDDM leaf images across 53 classes in a zero-shot setup. A non-participant GPT-5 judge scored reasoning quality on a 1-10 scale. The prompt supplied the crop and disease taxonomy but withheld the dataset name; output budget was approximately 600-700 tokens. Evaluation request date, temperature, and immutable serving snapshot are not reported.
- Context
- This is a task-specific zero-shot crop-disease VLM observation, not a general model-quality score. The reasoning rating comes from one independent judge model, not human raters. The study reports an alias string, not an immutable model snapshot or request timestamp. xAI currently documents this alias as grok-4.20-0309-non-reasoning, but the paper does not establish which snapshot served its earlier evaluation.
Page 1 of 2
Popularity compared
How often the model is searched for, used and mentioned, compared with every model of its category.
Popularity index
The 100 most popular models of this category, 0 to 100, higher is more popular
20 of 101 models
Gemini 4 Argon, Google DeepMind: 92
GPT-6 Astra, OpenAI: 78
Claude Haiku 5.5, Anthropic: 77
Qwen3.5 4B, Alibaba: 76
GPT-5, OpenAI: 73
GPT-6 Sol, OpenAI: 72
Qwen3.5 0.8B, Alibaba: 72
Qwen3.5 2B, Alibaba: 71
Qwen 3.8 27B, Alibaba: 70
Claude Opus 5.5, Anthropic: 70
DeepSeek V4.1 Flash 1017, DeepSeek: 68
Qwen 3.8 Max 0902, Alibaba: 68
Kimi K3, Moonshot AI: 68
DeepSeek-V4-Flash, DeepSeek: 65
GPT-5.6 Terra, OpenAI: 64
Claude Sonnet 5, Anthropic: 63
DiffusionGemma 26B A4B IT, Google: 63
Claude Opus 5, Anthropic: 62
Qwen 3.8 Flash, Alibaba: 62
Grok 4.20 Non-Reasoning, xAI: 0
As of November 2026 · Source: gradually.ai popularity index
Over time
Weekly values from October 18, 2026 to November 5, 2026
Grok 4.20 Non-ReasoningAgnes 3.0 FlashAion 3.5Aion 3.5 MiniEmber-1
All specifications
Input and 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
API
Prices
Not included in any subscription in Germany.
Frequently asked questions about Grok 4.20 Non-Reasoning
How large is the context window of Grok 4.20 Non-Reasoning?
1 M tokens. That is how much text Grok 4.20 Non-Reasoning reads at once, about 1,500 A4 pages.
Page count as a rule of thumb from Artificial Analysis
Grok 4.20 Non-ReasoningRank 31 of 1021M
Claude Opus 5.51M
Gemini 4 Carbon1M
Claude Fable 5.11M
Claude Sonnet 5.51M
Muse Wonder Full1M
Claude Opus 51M
How fast is Grok 4.20 Non-Reasoning?
88 tokens per second. That is faster than a typical model at 87.
Measured by Artificial Analysis on Nov 6, 2026. AA 3.0.0, including time to first token, averaged across available workloads and five days.
GPT-589
Grok 4.20 Non-ReasoningRank 48 of 10188
Claude Fable 5.188
GPT-5 mini87
Claude Fable 586
GPT-4.1 mini85
GLM-5.285
Can I run Grok 4.20 Non-Reasoning locally?
No. xAI does not publish the weights.
Stated by xAI
- Open weights
- no
- License
- n/a
- Parameters
- n/a
- active per token
- n/a
Sources and data date
Show all 18 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 | 2 | 0 | 2 |
Without vendor claims and historical or disputed results; those stay labelled in the single values.
| Type | Evidence and data date |
|---|---|
| Speed | SpaceXAI, Non-reasoning. AA 3.0.0, including time to first token, averaged across available workloads and five days. Retrieved November 6, 2026. |
| Research status | Research date October 4, 2026. 9 source URLs checked. This documents the inspected sources, not an exhaustive inventory of every publication. |
| Not published | benchmark-results Original wording The exact old Grok 4.20 0309 non-reasoning page is not a usable exact measured result: AA explicitly labels its composite estimate and says independent evaluation is forthcoming. The observed estimate (14.606641173969 index points) is retained only as a boundary note, not numeric evidence. The AA page for grok-4-20-non-reasoning is a separate 2026-04-07 v2 revision and was excluded. Reasoning and multi-agent records were not transferred. |
| Unresolved | official-quality-and-serving-revision Note The CDDM table reports measurements for the explicitly named grok-4.20-non-reasoning API route. Those measurements are not tied to immutable serving bytes, and additional official or Artificial Analysis scores for an exactly pinned revision remain undocumented. The AA estimate remains excluded as measured evidence. |
| Unresolved | serving-snapshot Note The paper names the exact grok-4.20-non-reasoning API string and describes standard production endpoints, but does not report the request date or immutable serving snapshot. xAI's current documentation maps the alias to grok-4.20-0309-non-reasoning; that current alias mapping does not prove which snapshot served the earlier paper run. |
| Unresolved | serving-snapshot Note The paper names the model/API string but does not publish the evaluation request date or immutable provider serving revision. |
| Additional source | xAI Grok 4.20 Non-Reasoning developer docs (retrieved October 3, 2026; October 4, 2026; October 15, 2026) · xAI Grok 4.20 Non-Reasoning developer docs · Editorial description reviewed October 4, 2026 |
| Additional source | Artificial Analysis (retrieved November 2, 2026) |
| Additional source | Artificial Analysis (retrieved November 2, 2026; November 6, 2026)12 |
| Additional source | Julien Talbot Hermes Grok Bench (retrieved October 15, 2026) |
| Additional source | Frontiers in Plant Science, VIDA+PANDA Tables 1 and 2 (retrieved October 16, 2026; October 18, 2026) |
| Additional source | UGI Leaderboard (retrieved October 21, 2026) |
| Additional source | Model metadata (retrieved October 15, 2026)12 |
| Additional source | Model metadata (retrieved October 16, 2026) |
| Additional source | OpenRouter: Popularity index (derived), CC BY 4.0 |
| Additional source | Vercel AI Gateway: Popularity index (derived), CC BY 4.0 |
| Additional source | Ollama: cumulative downloads for the popularity index |
| Additional source | Replicate: cumulative runs for the popularity index |
| Additional source | Google Trending Now: limited feed selection for Germany and the US |
| Additional source | Creative Commons Attribution 4.0 |