DeepSeek-V3.1 Terminus
Not tested yet
Language modelActive
DeepSeek-V3.1 Terminus
DeepSeek
- Released
- September 22, 2025
- Data date
- September 8, 2026
- Intelligence
- Overall score
- Price
- Input · Output / 1M tokens
No published category-wide ranking is available for this model.
Entries 1-12 of 14. 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.
MMLU-Pro (Show measurement, test conditions, and source)
- Source value
- 85
- Score
- 85
- Metric
- percent
- Unit
- %
- Category
- agentic
- Direction
- Higher is better
- Comparison cohort
- research:deepseek-v3.1-terminus:0
- Source type
- vendor-reported
- Evaluator
- DeepSeek V3.1-Terminus model card
- Status
- active
- Retrieved at
- 2026-09-08
- Methodology
- Vendor-reported result. reasoning and agent benchmark tables; source notes rolling/updated variant
GPQA (Show measurement, test conditions, and source)
- Source value
- 80.7
- Score
- 80.7
- Metric
- percent
- Unit
- %
- Category
- agentic
- Direction
- Higher is better
- Comparison cohort
- research:deepseek-v3.1-terminus:1
- Source type
- vendor-reported
- Evaluator
- DeepSeek V3.1-Terminus model card
- Status
- active
- Retrieved at
- 2026-09-08
- Methodology
- Vendor-reported result. reasoning and agent benchmark tables; source notes rolling/updated variant
HLE (Show measurement, test conditions, and source)
- Source value
- 21.7
- Score
- 21.7
- Metric
- percent
- Unit
- %
- Category
- agentic
- Direction
- Higher is better
- Comparison cohort
- research:deepseek-v3.1-terminus:2
- Source type
- vendor-reported
- Evaluator
- DeepSeek V3.1-Terminus model card
- Status
- active
- Retrieved at
- 2026-09-08
- Methodology
- Vendor-reported result. reasoning and agent benchmark tables; source notes rolling/updated variant
LiveCodeBench (Show measurement, test conditions, and source)
- Source value
- 74.9
- Score
- 74.9
- Metric
- percent
- Unit
- %
- Category
- agentic
- Direction
- Higher is better
- Comparison cohort
- research:deepseek-v3.1-terminus:3
- Source type
- vendor-reported
- Evaluator
- DeepSeek V3.1-Terminus model card
- Status
- active
- Retrieved at
- 2026-09-08
- Methodology
- Vendor-reported result. reasoning and agent benchmark tables; source notes rolling/updated variant
Aider-Polyglot (Show measurement, test conditions, and source)
- Source value
- 76.1
- Score
- 76.1
- Metric
- percent
- Unit
- %
- Category
- agentic
- Direction
- Higher is better
- Comparison cohort
- research:deepseek-v3.1-terminus:5
- Source type
- vendor-reported
- Evaluator
- DeepSeek V3.1-Terminus model card
- Status
- active
- Retrieved at
- 2026-09-08
- Methodology
- Vendor-reported result. reasoning and agent benchmark tables; source notes rolling/updated variant
BrowseComp (Show measurement, test conditions, and source)
- Source value
- 38.5
- Score
- 38.5
- Metric
- percent
- Unit
- %
- Category
- agentic
- Direction
- Higher is better
- Comparison cohort
- research:deepseek-v3.1-terminus:6
- Source type
- vendor-reported
- Evaluator
- DeepSeek V3.1-Terminus model card
- Status
- active
- Retrieved at
- 2026-09-08
- Methodology
- Vendor-reported result. reasoning and agent benchmark tables; source notes rolling/updated variant
BrowseComp-zh (Show measurement, test conditions, and source)
- Source value
- 45
- Score
- 45
- Metric
- percent
- Unit
- %
- Category
- agentic
- Direction
- Higher is better
- Comparison cohort
- research:deepseek-v3.1-terminus:7
- Source type
- vendor-reported
- Evaluator
- DeepSeek V3.1-Terminus model card
- Status
- active
- Retrieved at
- 2026-09-08
- Methodology
- Vendor-reported result. reasoning and agent benchmark tables; source notes rolling/updated variant
SimpleQA (Show measurement, test conditions, and source)
- Source value
- 96.8
- Score
- 96.8
- Metric
- percent
- Unit
- %
- Category
- agentic
- Direction
- Higher is better
- Comparison cohort
- research:deepseek-v3.1-terminus:8
- Source type
- vendor-reported
- Evaluator
- DeepSeek V3.1-Terminus model card
- Status
- active
- Retrieved at
- 2026-09-08
- Methodology
- Vendor-reported result. reasoning and agent benchmark tables; source notes rolling/updated variant
SWE Verified (Show measurement, test conditions, and source)
- Source value
- 68.4
- Score
- 68.4
- Metric
- percent
- Unit
- %
- Category
- agentic
- Direction
- Higher is better
- Comparison cohort
- research:deepseek-v3.1-terminus:9
- Source type
- vendor-reported
- Evaluator
- DeepSeek V3.1-Terminus model card
- Status
- active
- Retrieved at
- 2026-09-08
- Methodology
- Vendor-reported result. reasoning and agent benchmark tables; source notes rolling/updated variant
SWE multilingual (Show measurement, test conditions, and source)
- Source value
- 57.8
- Score
- 57.8
- Metric
- percent
- Unit
- %
- Category
- agentic
- Direction
- Higher is better
- Comparison cohort
- research:deepseek-v3.1-terminus:10
- Source type
- vendor-reported
- Evaluator
- DeepSeek V3.1-Terminus model card
- Status
- active
- Retrieved at
- 2026-09-08
- Methodology
- Vendor-reported result. reasoning and agent benchmark tables; source notes rolling/updated variant
Terminal-bench (Show measurement, test conditions, and source)
- Source value
- 36.7
- Score
- 36.7
- Metric
- percent
- Unit
- %
- Category
- agentic
- Direction
- Higher is better
- Comparison cohort
- research:deepseek-v3.1-terminus:11
- Source type
- vendor-reported
- Evaluator
- DeepSeek V3.1-Terminus model card
- Status
- active
- Retrieved at
- 2026-09-08
- Methodology
- Vendor-reported result. reasoning and agent benchmark tables; source notes rolling/updated variant
Codeforces rating (Show measurement, test conditions, and source)
- Source value
- 2046
- Score
- 2046
- Metric
- rating
- Unit
- rating
- Category
- agentic
- Direction
- Higher is better
- Comparison cohort
- research:deepseek-v3.1-terminus:4
- Source type
- vendor-reported
- Evaluator
- DeepSeek V3.1-Terminus model card
- Status
- active
- Retrieved at
- 2026-09-08
- Methodology
- Vendor-reported result. reasoning and agent benchmark tables; source notes rolling/updated variant
Our model tests
Community tests you can inspect
Compare how the models respond to the same prompt. Each test shows the first attempt, with no subsequent fixes to the generated code. These results do not contribute to an overall score.
Pelican on a bicycle
A community classic for free-form SVG drawing.
Prompt and test conditions
Generate an SVG of a pelican riding a bicycle
Token limit including reasoning: 8,192 tokens
Task origin (Simon Willison)Profile
Specifications and access
Published information about this model. Existing estimates are explicitly labeled.
Additional source observations
Dated documentation and model-card observations. Provider limits, native context and extended context can differ. Configuration notes retain the source wording.
- License
- MITSource
- Parameters
- 671 BSource
- Active parameters
- 37 BSource
- Rounded context label
- 128kSource
- Model configuration context
- 163,840 tokens (Native checkpoint context; distinct from the rounded model-card label and provider serving limits.)Source
- Architecture
- MoE hybrid reasoning modelSource
- Reasoning mode
- reasoning modeSource
- Tool calling
- YesSource
- Quantization
- FP8; known o_proj issue in released FP8 artifactSource
Pricing
Published prices
Prices apply to the stated unit. Resolution, output length, and provider can change the cost.
Cost example
100 API requests
$0.077
100 requests with 1,000 input and 500 output tokens each calculate to $0.077. Input accounts for $0.027; output accounts for $0.05.
- Input price
- $0.27 / 1M tokensBase tier for each request
- Output price
- $1 / 1M tokensBase tier for each request
The calculation uses documented token prices and no cache discount. It excludes extra tools, tax, reasoning tokens, and further hidden output tokens.
Open sourceHead-to-head comparisons
Compare this model
Each matchup compares this model with exactly one other model from the same category.
There are no published direct comparisons for this model yet.
More models
Models from the same selection
All AI models
Evidence
Sources and data date
Every statement links to its underlying documentation or leaderboard.
- Price
Base API prices without caching or batch discounts. Higher context tiers and other rates are listed under Costs.
Research date September 8, 2026. 4 source URLs checked. This documents the inspected sources, not an exhaustive inventory of every publication.
Open evidence questions and research notes (original wording)
Unresolved
release-date. No exact release-date field in card.
- Model metadataSource
- API pricingSource
- DeepSeek V3.1-Terminus model card (retrieved September 8, 2026)Source
- Artificial Analysis (retrieved September 8, 2026)Source
- Model metadata (retrieved September 8, 2026)Source
- Model metadata (retrieved September 8, 2026)Source
- DeepSeek-V3.2-Exp announcement (2026-09-09)Source