The LLM benchmark aggregator.

DeepSeek V3.1 Terminus

DeepSeek · 685B · open weights

Strongest in Core abilities (#77 of 204, on 1 of its 3 boards), weakest in Agents (#160 of 267). Above par in 4 of 6 scopes. Among the models it meets almost everywhere, it finishes behind Claude Opus 5.5 and Claude Fable 5 and ahead of Granite 4.2 30B and Qwen3.5 122B A10B.

§ 1 · Profile

What it is good at

Bars run from 50 — the average of the models each source lists — so right of the line is above par. The middle column is the gap to whoever leads that scope.

Core abilities51.6−15.8#77/2041/3

General intelligence52.3−17.8#78/2041/3

Human preference58.6−8.9#104/3411/1

Agents46.2−21.9#160/2672/5

Knowledge work45.1−28.5#105/1772/2

§ 2 · Head to head

What it beats, and what beats it

The same models turn up scope after scope. Counted once: where both were placed, who finished higher. Bars run right for DeepSeek V3.1 Terminus, left for the other.

K2 Horizon 0.9Bwon 4 · lost 0

Ministral 3 3Bwon 4 · lost 0

Ministral 3 8Bwon 4 · lost 0

Ministral 3 14Bwon 4 · lost 0

Celeris 1won 4 · lost 0

Mistral Small 4won 4 · lost 0

Qwen3 Coder Nextwon 4 · lost 0

North Mini Codewon 4 · lost 0

MiniCPM5 2Bwon 4 · lost 0

Mercury 2.5won 4 · lost 0

Nemotron 3.5 Lightningwon 4 · lost 0

Command A+won 4 · lost 0

Qwen3 14Bwon 4 · lost 0

Nemotron 3 Ultrawon 4 · lost 0

Qwen3 30B A3B Thinkingwon 4 · lost 0

Llama 3.1 8B Instructwon 4 · lost 0

Claude 3 Haikuwon 4 · lost 0

GPT-4won 4 · lost 0

Llama 3 70B Instructwon 4 · lost 0

Magistral Mediumwon 4 · lost 0

Mercurywon 4 · lost 0

Gemma 3n E4b Itwon 4 · lost 0

Llama 3.3 70B Instructwon 4 · lost 0

Claude 3.5 Haikuwon 4 · lost 0

Nova 2 Litewon 4 · lost 0

Gemma 3 12B Itwon 4 · lost 0

Mercury 2won 4 · lost 0

Gemini 2.0 Flashwon 4 · lost 0

O3 Miniwon 4 · lost 0

GLM-4.5 Airwon 4 · lost 0

MiniMax M2.5won 4 · lost 0

Qwen3 Next 80B A3B Instructwon 4 · lost 0

Qwen3 235B A22B Thinkingwon 4 · lost 0

DeepSeek V3.1won 4 · lost 0

Gemma 3 4B Itwon 5 · lost 0

Granite 4.2 8Bwon 6 · lost 0

Granite 4.2 3Bwon 6 · lost 0

GPT-4o Miniwon 6 · lost 0

GPT OSS 20Bwon 6 · lost 0

Llama 4 Scout Instructwon 6 · lost 0

GPT-4.1 Nanowon 6 · lost 0

Llama 4 Maverick Instructwon 6 · lost 0

Gemma 3 27B Itwon 6 · lost 0

Trinity Largewon 6 · lost 0

DeepSeek V3won 6 · lost 0

Mistral Large 3won 6 · lost 0

Qwen3 32Bwon 5 · lost 1

GPT-4.1 Miniwon 5 · lost 1

Command Awon 4 · lost 1

Mistral Smallwon 4 · lost 1

Gemini 2.5 Flash Litewon 4 · lost 1

Mistral Mediumwon 4 · lost 1

Gemini 2.5 Flashwon 4 · lost 1

GPT-4.1won 4 · lost 1

GPT OSS 120Bwon 4 · lost 2

Qwen3 Next 80B A3B Thinkingwon 4 · lost 2

GPT-5.4 Nanowon 4 · lost 2

Claude Haiku 4.5won 4 · lost 2

Qwen3.5 122B A10Bwon 4 · lost 2

Granite 4.2 30Bwon 3 · lost 2

Qwen3.6 27Bwon 2 · lost 2

Kimi K2.7 Codewon 2 · lost 2

Nemotron 3 Ultra 550B A55Bwon 2 · lost 2

Grok Build 0.1won 2 · lost 2

K2 Horizon 3.7Bwon 2 · lost 2

Ring 2.6.1Twon 2 · lost 2

Qwen3.6 35B A3Bwon 2 · lost 2

LongCat 2.0won 2 · lost 2

K2 Horizon 7Bwon 2 · lost 2

Ling 3.0 Flash Finwon 2 · lost 2

Ling 3.0 Flash VLwon 2 · lost 2

K2 Horizon MoVA 36B A4Bwon 2 · lost 2

Apodex 1.1won 2 · lost 2

MiniMax M2won 2 · lost 2

Qwen Pluswon 2 · lost 2

Step 3won 2 · lost 2

GLM-4.5vwon 2 · lost 2

NVIDIA Nemotron 3 Super 120B A12Bwon 2 · lost 2

Grok 3 Miniwon 2 · lost 2

Mimo V2 Flashwon 2 · lost 2

Qwen3.5 35B A3Bwon 2 · lost 2

Grok 4 Fastwon 2 · lost 2

Inkling Smallwon 2 · lost 2

GLM-4.5won 2 · lost 2

Grok 3won 2 · lost 2

MiniMax M2.7won 2 · lost 2

Longcat Flash Chatwon 2 · lost 2

GPT-5 Nanowon 2 · lost 3

O4 Miniwon 2 · lost 3

Grok 4won 2 · lost 3

Kimi K2won 2 · lost 3

GPT-5 Miniwon 2 · lost 4

Claude Sonnet 4won 2 · lost 4

DeepSeek R1won 2 · lost 4

Muse Glimmerwon 2 · lost 4

GLM-4.6won 2 · lost 4

DeepSeek V3.2won 2 · lost 4

Mistral Medium 3.5won 2 · lost 4

Gemini 3.1 Flash Litewon 2 · lost 4

DeepSeek V4 Flashwon 2 · lost 4

MiniMax M3won 2 · lost 4

GPT-6 Lunawon 2 · lost 4

Inklingwon 2 · lost 4

Grok 4.3won 2 · lost 4

Qwen3.6 Pluswon 2 · lost 4

Gemini 2.5 Prowon 2 · lost 4

GPT-5.4 Miniwon 2 · lost 4

Gemma 4 31Bwon 2 · lost 4

GPT-5.6 Lunawon 2 · lost 4

Gemini 3.5 Flash Litewon 2 · lost 4

Qwen3.7 Pluswon 2 · lost 4

Hy3won 2 · lost 4

Kimi K2.6won 2 · lost 4

Mimo V2.5 Prowon 2 · lost 4

Claude Sonnet 4.6won 2 · lost 4

GPT-5.5 Instantwon 2 · lost 4

Qwen3.7 Maxwon 2 · lost 4

GLM-5.2won 2 · lost 4

Gemini 3.6 Flashwon 2 · lost 4

Muse Spark 1.1won 2 · lost 4

Claude Sonnet 5won 1 · lost 5

GLM-5.3 Flashwon 1 · lost 5

Gemini 3.5 Flashwon 1 · lost 5

K2 Horizon 375B A23Bwon 0 · lost 4

Qwen3.8 Flash Nextwon 0 · lost 4

DeepSeek V4 Flash Visionwon 0 · lost 4

Qwen3.8 2.4T A95Bwon 0 · lost 4

Step 5won 0 · lost 4

Qwen3 Maxwon 0 · lost 4

Claude Opus 4won 0 · lost 4

Mimo V2.5won 0 · lost 4

Gemma 4 26B A4Bwon 0 · lost 4

GLM-4.7won 0 · lost 4

ChatGPT-4owon 0 · lost 4

GPT-5.1won 0 · lost 4

GLM-5won 0 · lost 4

GLM-5.1won 0 · lost 4

Grok 4.20won 0 · lost 4

Muse Sparkwon 0 · lost 4

Claude Opus 4.6won 0 · lost 4

Qwen3 235B A22B Instructwon 0 · lost 5

O3won 0 · lost 5

GPT-5.2won 0 · lost 5

Claude Opus 4.5won 0 · lost 5

Gemini 3 Prowon 0 · lost 5

GPT-5won 0 · lost 6

Qwen3.8 27Bwon 0 · lost 6

Grok 4.7won 0 · lost 6

Qwen3.5 397B A17Bwon 0 · lost 6

Kimi K2.5won 0 · lost 6

Grok 4.6won 0 · lost 6

GPT-6 Solwon 0 · lost 6

Claude Sonnet 4.5won 0 · lost 6

DeepSeek V4 Prowon 0 · lost 6

Grok 4.5won 0 · lost 6

GPT-5.6 Terrawon 0 · lost 6

GPT-5.4won 0 · lost 6

DeepSeek V4.1 Flashwon 0 · lost 6

GPT-6 Astrawon 0 · lost 6

Qwen3.8 Maxwon 0 · lost 6

Mimo V2.6 Prowon 0 · lost 6

GLM-5.3won 0 · lost 6

Claude Opus 4.8won 0 · lost 6

GPT-5.5won 0 · lost 6

GPT-5.6 Solwon 0 · lost 6

Gemini 3.1 Prowon 0 · lost 6

Gemini 3.7 Flashwon 0 · lost 6

Kimi K3won 0 · lost 6

Claude Opus 5won 0 · lost 6

Gemini 3.8 Flashwon 0 · lost 6

Muse Spark 1.3won 0 · lost 6

Muse Spark 1.2won 0 · lost 6

Claude Fable 5.1won 0 · lost 6

Claude Opus 4.7won 0 · lost 6

Claude Fable 5won 0 · lost 6

Claude Opus 5.5won 0 · lost 6

§ 3 · Sources

Where the numbers come from

4 publications, 4 figures. Every one links to the page it was read from.

LMArena Text ↗1 measureread 2026-09-27

LMArena Text 1417

GDPval-AA ↗1 measureread 2026-09-27

GDPval-AA 698

AA-Briefcase ↗1 measureread 2026-09-27

AA-Briefcase 457

Kagi LLM Benchmark ↗1 measureread 2026-09-27

Kagi LLM Benchmark 57.4%

Badge

[![PublicAI Index](https://publicai.io/model-index/badge?model=deepseek-v3-1-terminus)](https://publicai.io/model-index/m/deepseek-v3-1-terminus)