Skip to content

DeepSeek · Chat

DeepSeek V3.2

DeepSeek 🇨🇳ChatDeepSeek familyOpen weights. One of its sources says the weights are publicly available.The maker currently lists 1 Jul 2027 as the deprecation date. Plans can change.

Newer version: DeepSeek V4. Our catalogue has no release date for DeepSeek V4. We compared the version names in our catalogue.

From Venice's listing:
DeepSeek-V3.2 is an efficient large language model with DeepSeek Sparse Attention (DSA) for long contexts. It features strong reasoning and tool-use skills, achieving top results on the 2025 IMO and IOI.
llminfo IQ
125.0
Max context
164K
Listings
78

Evidence and sources

🤗 deepseek-ai/DeepSeek-V3.2 Model architecture, as its sources state it llminfo IQ 125.0, ranked among chat models. From 247 benchmarks. Plenty of data behind it.How llminfo IQ is computedCoding IQ 101.2, 95% interval 94–108. From 11 benchmark families across 9 sources. Plenty of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ.How Coding IQ is computedGood value (63.8). Cheaper per task than its benchmark results predict. From 42 benchmarks with measured task costs. Plenty of data behind it.How Value for money worksReleased 1 Dec 2025Maker HQ: CN (according to DeepSeek's own website)
All rankings: #36 on NanoGPT recently #54 on OpenRouter today #61 on OpenRouter this month #65 on OpenRouter this weekSources: LMArena data, AWS Bedrock Runtime data, LiteLLM data, Models.dev data, Morpheus Inference Marketplace data, ZenMux data, Venice.ai data, Hugging FacePrimary data from LMArena from AWS Bedrock Runtime: name from LiteLLM: deprecation date from Models.dev: weights availability from Morpheus Inference Marketplace: architecture from ZenMux: primary mode from Venice.ai: description from Hugging Face: release date
Tags from its sources:instructtoolsreasoningopen-sourcelong-contextcoding

Advantages and trade-offs

Compared with 10 chat models, picked mainly by popularity and IQ
We only make a claim when there's enough data on models of the same type. Updated 22 Nov, 06:07 UTC.Prices come from the model's maker where they sell it, otherwise their official partner, otherwise the market median. This comparison uses 9 maker's prices; 1 market median price.

Input & output

as stated by the model's providers
Text
Text

Capabilities

Supported operations and provider-stated features

ReasoningFunction callingStructured outputPrompt cachingWeb searchBatch API

Listings

 DeepSeekdeepseek-v3.2Sources: GenAI Prices data, LiteLLM data164K$0.2288cache read $0.028$0.3432———
 Tokengodeepseek/deepseek-v3.2Source: Models.dev data128K$0.2174cache read $0.06$0.326———
 Aviandeepseek/deepseek-v3.2Source: Avian.io Models data164K$0.23cache read $0.012$0.33———
 OpenRouterInference by SiliconFlowdeepseek/deepseek-v3.2 · fp8Source: OpenRouter data docs164K$0.259cache read $0.14$0.80—171,972.5ms
 Atlasclouddeepseek-ai/deepseek-v3.2Source: AtlasCloud data164K$0.26cache read $0.13$0.38———
 Fastrouterdeepseek/deepseek-v3.2Source: FastRouter data164K$0.26cache read $0.13$0.38———
 Flowdotdeepseek-ai/DeepSeek-V3.2Source: FlowDot data164K$0.26$0.38———
 Hugging FaceInference by DeepInfradeepseek-ai/DeepSeek-V3.2Source: Hugging Face Inference Providers data164K$0.26$0.38———
 Kilodeepseek/deepseek-v3.2Sources: Kilo AI Gateway data, Models.dev data164K$0.26cache read $0.13$0.38$0.00——
 LlmgatewayInference by DeepInfradeepseek-ai/DeepSeek-V3.2Source: LLMGateway dataNot published by the provider$0.26cache read $0.13$0.38———

6 listings without a price are hidden.

Included in plans

BytePlus Coding Plan LiteBytePlusSubscription$10/moPlan page
BytePlus Coding Plan ProBytePlusSubscription$50/moPlan page

Benchmarks

Named effort levels run from lowest to highest. Select a result to see its source details.

Human preference

Arena ELO51 related benchmarks · 101 results
Source: LMArena data · Nov 2026
Arena ELO
Chinese
Creative Writing
English
Exclude Ties
Expert
Show 89 more resultsHide 89 results
French
German
Hard Prompts
Hard Prompts English
Industry Entertainment and Sports and Media
Industry Life and Physical and Social Science
Industry Mathematical
Industry Medicine and Healthcare
Industry Software and It Services
Industry Writing and Literature and Language
Japanese
Korean
Longer Query
Multi Turn
Non English
Polish
Russian
Spanish
Style-Controlled
Text Factuality
Text Factuality - Chinese
Text Factuality - Creative Writing
Text Factuality - English
Text Factuality - Exclude Ties
Text Factuality - Expert
Text Factuality - French
Text Factuality - German
Text Factuality - Hard Prompts
Text Factuality - Hard Prompts English
Text Factuality - Industry Mathematical
Text Factuality - Longer Query
Text Factuality - Multi Turn
Text Factuality - Non English
Text Factuality - Polish
Text Factuality - Russian
Text Factuality - Spanish
Design Arena EloSource: Design Arena Leaderboards data · Nov 2026

Agentic

BrowseComp5 results
Reported by DeepSeek · via BenchBoard docs · Reported by the model's developer. The source relaying it did not verify it.
Thinking mode · temperature 1.0 · 128K context
Thinking mode · temperature 1.0 · 128K context · commercial search API · baseline · Section 4.4
Thinking mode · temperature 1.0 · 128K context · commercial search API · without context management · Table 2
Source: LLM Stats data · Reported by the model's developer. The source relaying it did not verify it. claim
Setting not stated
BrowseComp-zh2 results
Reported by DeepSeek · via BenchBoard docs · Reported by the model's developer. The source relaying it did not verify it.
Thinking mode · temperature 1.0 · 128K context · commercial search API
Source: LLM Stats data · Reported by the model's developer. The source relaying it did not verify it. claim
Setting not stated
GAIA4 results
Source: GAIA Leaderboard data · Jan 2026
Difficulty level
Difficulty not reported
MCP-MarkSource: LLM Stats data · Reported by the model's developer. The source relaying it did not verify it. claim
MCP-Universe3 results
Reported by DeepSeek · via BenchBoard docs · Reported by the model's developer. The source relaying it did not verify it.
temperature 1.0 · 128K context · internal search/Playwright environment · tool-role outputs
Source: LLM Stats data · Reported by the model's developer. The source relaying it did not verify it. claim
Setting not stated
Open Agent Leaderboard6 related benchmarks · 30 results
SWE-bench
Source: Open Agent Leaderboard results data docs
By agent · task: swebench · route: openai/azure
AppWorld test_normal
Source: Open Agent Leaderboard results data docs
By agent · task: appworld_test_normal · subset: test_normal · route: openai/azure
BrowseCompPlus
Source: Open Agent Leaderboard results data docs
By agent · task: browsecompplus · route: openai/azure
Tau Bench 2 airline
Source: Open Agent Leaderboard results data docs
By agent · task: tau2_airline · subset: airline · route: openai/azure
Tau Bench 2 retail
Source: Open Agent Leaderboard results data docs
By agent · task: tau2_retail · subset: retail · route: openai/azure
Tau Bench 2 telecom
Source: Open Agent Leaderboard results data docs
By agent · task: tau2_telecom · subset: telecom · route: openai/azure
tau2-bench4 related benchmarks · 14 results
airline
Source: OpenRouter benchmark exports
1 Dec 2025 version · 90-day results
Across 10 providers lowest 59.2% median 72.1% highest 74.1%OpenRouter's automatic routing 73.3%
Show results from 10 providers and OpenRouter's automatic routingHide results from 10 providers and OpenRouter's automatic routing
  • Venice
  • DeepInfra
  • OpenRouter's automatic routing
  • SiliconFlow
  • StreamLake
  • Friendli
  • Phala
  • Alibaba
  • GMICloud
  • Google
  • AtlasCloud
v0.1 airline pass^1
Source: Sierra tau-bench Leaderboard data docs · Dec 2025
submission-deepseek-v3.2_deepseek_2025-12-01, usersim-deepseek-v3.2
v0.1 retail pass^1
Source: Sierra tau-bench Leaderboard data docs · Dec 2025
submission-deepseek-v3.2_deepseek_2025-12-01, usersim-deepseek-v3.2
v0.1 telecom pass^1
Source: Sierra tau-bench Leaderboard data docs · Dec 2025
submission-deepseek-v3.2_deepseek_2025-12-01, usersim-deepseek-v3.2
Terminal-Bench3 related benchmarks · 5 results
2.0 · Terminus-2
Reported by DeepSeek · via BenchBoard docs · Reported by the model's developer. The source relaying it did not verify it.
temperature 1.0 · 128K context · Claude Code framework
Non-thinking mode · temperature 1.0 · 128K context · Terminus framework
Terminal-Bench 2.0
Source: LLM Stats data · Reported by the model's developer. The source relaying it did not verify it. claim

Reasoning

CL-bench9 related benchmarks · 27 results
Overall
Reported by Tencent Hunyuan / Fudan NLP Group · via BenchBoard docs
By site model-mode label · benchmark-owner leaderboard run · 1899 tasks · GPT-5.1 judge · low reasoning effort · candidate API snapshot and exact candidate harness not specified on public leaderboard
Source: Epoch AI Benchmarking Hub data
Setting not stated
Domain Knowledge Reasoning
Reported by Tencent Hunyuan / Fudan NLP Group · via BenchBoard docs
By site model-mode label · benchmark-owner leaderboard run · 1899 tasks · GPT-5.1 judge · low reasoning effort · candidate API snapshot and exact candidate harness not specified on public leaderboard
Source: Epoch AI Benchmarking Hub data
Setting not stated
Empirical Discovery & Simulation
Reported by Tencent Hunyuan / Fudan NLP Group · via BenchBoard docs
By site model-mode label · benchmark-owner leaderboard run · 1899 tasks · GPT-5.1 judge · low reasoning effort · candidate API snapshot and exact candidate harness not specified on public leaderboard
Source: Epoch AI Benchmarking Hub data
Setting not stated
Life · Behavioral Records & Activity Trails
Reported by Tencent Hunyuan / Fudan NLP Group · via BenchBoard docs
benchmark-owner leaderboard run · site model-mode label: DeepSeek V3.2 · 405 tasks · GPT-5.1 judge · high reasoning effort · single run; no standard deviation · candidate API snapshot and exact candidate harness not specified on public leaderboard
benchmark-owner leaderboard run · site model-mode label: DeepSeek V3.2 Thinking · 405 tasks · GPT-5.1 judge · high reasoning effort · candidate API snapshot and exact candidate harness not specified on public leaderboard
Source: Epoch AI Benchmarking Hub data
Setting not stated
Life · Communication & Social Interactions
Reported by Tencent Hunyuan / Fudan NLP Group · via BenchBoard docs
benchmark-owner leaderboard run · site model-mode label: DeepSeek V3.2 · 405 tasks · GPT-5.1 judge · high reasoning effort · single run; no standard deviation · candidate API snapshot and exact candidate harness not specified on public leaderboard
benchmark-owner leaderboard run · site model-mode label: DeepSeek V3.2 Thinking · 405 tasks · GPT-5.1 judge · high reasoning effort · candidate API snapshot and exact candidate harness not specified on public leaderboard
Source: Epoch AI Benchmarking Hub data
Setting not stated
Life · Fragmented Information & Revisions
Reported by Tencent Hunyuan / Fudan NLP Group · via BenchBoard docs
benchmark-owner leaderboard run · site model-mode label: DeepSeek V3.2 · 405 tasks · GPT-5.1 judge · high reasoning effort · single run; no standard deviation · candidate API snapshot and exact candidate harness not specified on public leaderboard
benchmark-owner leaderboard run · site model-mode label: DeepSeek V3.2 Thinking · 405 tasks · GPT-5.1 judge · high reasoning effort · candidate API snapshot and exact candidate harness not specified on public leaderboard
Source: Epoch AI Benchmarking Hub data
Setting not stated
Show 9 more resultsHide 9 results
Life · Overall
Reported by Tencent Hunyuan / Fudan NLP Group · via BenchBoard docs
benchmark-owner leaderboard run · site model-mode label: DeepSeek V3.2 · 405 tasks · GPT-5.1 judge · high reasoning effort · single run; no standard deviation · candidate API snapshot and exact candidate harness not specified on public leaderboard
benchmark-owner leaderboard run · site model-mode label: DeepSeek V3.2 Thinking · 405 tasks · GPT-5.1 judge · high reasoning effort · candidate API snapshot and exact candidate harness not specified on public leaderboard
Source: Epoch AI Benchmarking Hub data
Setting not stated
Procedural Task Execution
Reported by Tencent Hunyuan / Fudan NLP Group · via BenchBoard docs
By site model-mode label · benchmark-owner leaderboard run · 1899 tasks · GPT-5.1 judge · low reasoning effort · candidate API snapshot and exact candidate harness not specified on public leaderboard
Source: Epoch AI Benchmarking Hub data
Setting not stated
Rule System Application
Reported by Tencent Hunyuan / Fudan NLP Group · via BenchBoard docs
By site model-mode label · benchmark-owner leaderboard run · 1899 tasks · GPT-5.1 judge · low reasoning effort · candidate API snapshot and exact candidate harness not specified on public leaderboard
Source: Epoch AI Benchmarking Hub data
Setting not stated
Reported by DeepSeek · via BenchBoard docs · Reported by the model's developer. The source relaying it did not verify it.
Thinking mode · temperature 1.0 · 128K context · custom step-by-step / boxed-answer prompt
Thinking mode · temperature 1.0 · 128K context · official HLE template
Source: LLM Stats data · Reported by the model's developer. The source relaying it did not verify it. claim
Setting not stated
t2-benchSource: LLM Stats data · Reported by the model's developer. The source relaying it did not verify it. claim
Reported by DeepSeek · via BenchBoard docs · Reported by the model's developer. The source relaying it did not verify it.
Non-thinking mode · temperature 1.0 · 128K context · pass@1
Thinking mode · temperature 1.0 · 128K context · standard function-call format

Coding

Codeforces RatingThinking mode · temperature 1.0 · 128K context · reported rating · Reported by DeepSeek · via BenchBoard docs · Reported by the model's developer. The source relaying it did not verify it.
LiveCodeBench2 results
Reported by DeepSeek · via BenchBoard docs · Reported by the model's developer. The source relaying it did not verify it.
Thinking mode · temperature 1.0 · 128K context · pass@1-CoT
Source: LLM Stats data · Reported by the model's developer. The source relaying it did not verify it. claim
Setting not stated
SWE-bench Verified6 related benchmarks · 10 results
SWE-bench Verified
Reported by DeepSeek · via BenchBoard docs · Reported by the model's developer. The source relaying it did not verify it.
Non-thinking mode · temperature 1.0 · 128K context · resolved
Thinking mode · temperature 1.0 · 128K context · DeepSeek internal code-agent framework
Source: LLM Stats data · Reported by the model's developer. The source relaying it did not verify it. claim
Setting not stated
SWE-bench Multilingual
Reported by DeepSeek · via BenchBoard docs · Reported by the model's developer. The source relaying it did not verify it.
temperature 1.0 · 128K context · resolved
Source: LLM Stats data · Reported by the model's developer. The source relaying it did not verify it. claim
Setting not stated
ValsSource: Vals AI

Science

GPQASource: LLM Stats data · Reported by the model's developer. The source relaying it did not verify it. claim
GPQA Diamond18 results
Reported by DeepSeek · via BenchBoard docs · Reported by the model's developer. The source relaying it did not verify it.
Thinking mode · temperature 1.0 · 128K context · pass@1
Source: OpenEvals Leaderboard data · Reported by the model's developer. The source relaying it did not verify it. · Mar 2026
Setting not stated
Source: OpenRouter benchmark exports
1 Dec 2025 version · 90-day results
Across 15 providers lowest 59.3% median 81.7% highest 83.2%OpenRouter's automatic routing 82.4%
Show results from 15 providers and OpenRouter's automatic routingHide results from 15 providers and OpenRouter's automatic routing
  • AtlasCloud
  • Google
  • DigitalOcean
  • Alibaba
  • OpenRouter's automatic routing
  • Baidu
  • Venice
  • Phala
  • Friendli
  • SiliconFlow
  • StreamLake
  • GMICloud
  • Novita
  • DeepInfra
  • Mara
  • SambaNova

Math

AIME 20252 results
Reported by DeepSeek · via BenchBoard docs · Reported by the model's developer. The source relaying it did not verify it.
Thinking mode · temperature 1.0 · 128K context · custom step-by-step / boxed-answer prompt
Source: LLM Stats data · Reported by the model's developer. The source relaying it did not verify it. claim
Setting not stated
HMMT4 related benchmarks · 4 results
HMMT 2025
Source: LLM Stats data · Reported by the model's developer. The source relaying it did not verify it. claim
Feb 25
Reported by DeepSeek · via BenchBoard docs · Reported by the model's developer. The source relaying it did not verify it.
Thinking mode · temperature 1.0 · 128K context · custom step-by-step / boxed-answer prompt
February 2026
Source: OpenEvals Leaderboard data · Reported by the model's developer. The source relaying it did not verify it. · Mar 2026
Nov 25
Reported by DeepSeek · via BenchBoard docs · Reported by the model's developer. The source relaying it did not verify it.
Thinking mode · temperature 1.0 · 128K context · custom step-by-step / boxed-answer prompt
Reported by DeepSeek · via BenchBoard docs · Reported by the model's developer. The source relaying it did not verify it.
Thinking mode · temperature 1.0 · 128K context · custom step-by-step / boxed-answer prompt
Source: LLM Stats data · Reported by the model's developer. The source relaying it did not verify it. claim
Setting not stated

Multilingual

EKI EstonianSource: Keelemudelite mõõdupuu data docs

Vision

General

Instruction following

Knowledge

MMLU-Pro2 results
Reported by DeepSeek · via BenchBoard docs · Reported by the model's developer. The source relaying it did not verify it.
Thinking mode · temperature 1.0 · 128K context · exact match
Source: LLM Stats data · Reported by the model's developer. The source relaying it did not verify it. claim
Setting not stated

Factuality

Other

Claw-Eval v1.1Reported by Claw-Eval · via BenchBoard docs
MCPMark · Version Unspecifiedtemperature 1.0 · 128K context · internal search/Playwright environment · tool-role outputs · Reported by DeepSeek · via BenchBoard docs · Reported by the model's developer. The source relaying it did not verify it.
Nejumi 428 related benchmarks · 28 results
ALT_バイアス
Source: Nejumi Leaderboard 4 data docs
deepseek-ai/DeepSeek-V3.2: reasoning-enabled
ALT_倫理・道徳
Source: Nejumi Leaderboard 4 data docs
deepseek-ai/DeepSeek-V3.2: reasoning-enabled
ALT_制御性
Source: Nejumi Leaderboard 4 data docs
deepseek-ai/DeepSeek-V3.2: reasoning-enabled
ALT_堅牢性
Source: Nejumi Leaderboard 4 data docs
deepseek-ai/DeepSeek-V3.2: reasoning-enabled
ALT_毒性
Source: Nejumi Leaderboard 4 data docs
deepseek-ai/DeepSeek-V3.2: reasoning-enabled
ALT_真実性
Source: Nejumi Leaderboard 4 data docs
deepseek-ai/DeepSeek-V3.2: reasoning-enabled
Show 22 more resultsHide 22 results
AVG_jaster_0shot
Source: Nejumi Leaderboard 4 data docs
deepseek-ai/DeepSeek-V3.2: reasoning-enabled
AVG_jaster_2shots
Source: Nejumi Leaderboard 4 data docs
deepseek-ai/DeepSeek-V3.2: reasoning-enabled
GLP_アプリケーション開発
Source: Nejumi Leaderboard 4 data docs
deepseek-ai/DeepSeek-V3.2: reasoning-enabled
GLP_コーディング
Source: Nejumi Leaderboard 4 data docs
deepseek-ai/DeepSeek-V3.2: reasoning-enabled
GLP_一般的知識
Source: Nejumi Leaderboard 4 data docs
deepseek-ai/DeepSeek-V3.2: reasoning-enabled
GLP_基礎的言語性能
Source: Nejumi Leaderboard 4 data docs
deepseek-ai/DeepSeek-V3.2: reasoning-enabled
GLP_専門的知識
Source: Nejumi Leaderboard 4 data docs
deepseek-ai/DeepSeek-V3.2: reasoning-enabled
GLP_応用的言語性能
Source: Nejumi Leaderboard 4 data docs
deepseek-ai/DeepSeek-V3.2: reasoning-enabled
GLP_情報検索
Source: Nejumi Leaderboard 4 data docs
deepseek-ai/DeepSeek-V3.2: reasoning-enabled
GLP_意味解析
Source: Nejumi Leaderboard 4 data docs
deepseek-ai/DeepSeek-V3.2: reasoning-enabled
GLP_抽象的推論
Source: Nejumi Leaderboard 4 data docs
deepseek-ai/DeepSeek-V3.2: reasoning-enabled
GLP_推論能力
Source: Nejumi Leaderboard 4 data docs
deepseek-ai/DeepSeek-V3.2: reasoning-enabled
GLP_数学的推論
Source: Nejumi Leaderboard 4 data docs
deepseek-ai/DeepSeek-V3.2: reasoning-enabled
GLP_構文解析
Source: Nejumi Leaderboard 4 data docs
deepseek-ai/DeepSeek-V3.2: reasoning-enabled
GLP_知識・質問応答
Source: Nejumi Leaderboard 4 data docs
deepseek-ai/DeepSeek-V3.2: reasoning-enabled
GLP_翻訳
Source: Nejumi Leaderboard 4 data docs
deepseek-ai/DeepSeek-V3.2: reasoning-enabled
GLP_表現
Source: Nejumi Leaderboard 4 data docs
deepseek-ai/DeepSeek-V3.2: reasoning-enabled
GLP_論理的推論
Source: Nejumi Leaderboard 4 data docs
deepseek-ai/DeepSeek-V3.2: reasoning-enabled
GLP_関数呼び出し
Source: Nejumi Leaderboard 4 data docs
deepseek-ai/DeepSeek-V3.2: reasoning-enabled
TOTAL_SCORE
Source: Nejumi Leaderboard 4 data docs
deepseek-ai/DeepSeek-V3.2: reasoning-enabled
アラインメント(ALT)_AVG
Source: Nejumi Leaderboard 4 data docs
deepseek-ai/DeepSeek-V3.2: reasoning-enabled
汎用的言語性能(GLP)_AVG
Source: Nejumi Leaderboard 4 data docs
deepseek-ai/DeepSeek-V3.2: reasoning-enabled
SurveyReview v1.012 related benchmarks · 12 results
Structure · MAE
Reported by SurveyReview authors / Z.ai · via BenchBoard docs
SurveyReview v1.0 published leaderboard · text survey-paper evaluation · exact candidate snapshot, reasoning effort, prompt, run count, and harness are not specified in the summary table
Comprehensiveness · MAE
Reported by SurveyReview authors / Z.ai · via BenchBoard docs
SurveyReview v1.0 published leaderboard · text survey-paper evaluation · exact candidate snapshot, reasoning effort, prompt, run count, and harness are not specified in the summary table
Comprehensiveness · MSE
Reported by SurveyReview authors / Z.ai · via BenchBoard docs
SurveyReview v1.0 published leaderboard · text survey-paper evaluation · exact candidate snapshot, reasoning effort, prompt, run count, and harness are not specified in the summary table
Criticalness · MAE
Reported by SurveyReview authors / Z.ai · via BenchBoard docs
SurveyReview v1.0 published leaderboard · text survey-paper evaluation · exact candidate snapshot, reasoning effort, prompt, run count, and harness are not specified in the summary table
Criticalness · MSE
Reported by SurveyReview authors / Z.ai · via BenchBoard docs
SurveyReview v1.0 published leaderboard · text survey-paper evaluation · exact candidate snapshot, reasoning effort, prompt, run count, and harness are not specified in the summary table
Four dimensions · Average MAE
Reported by SurveyReview authors / Z.ai · via BenchBoard docs
SurveyReview v1.0 published leaderboard · text survey-paper evaluation · exact candidate snapshot, reasoning effort, prompt, run count, and harness are not specified in the summary table
Show 6 more resultsHide 6 results
Four dimensions · Average MSE
Reported by SurveyReview authors / Z.ai · via BenchBoard docs
SurveyReview v1.0 published leaderboard · text survey-paper evaluation · exact candidate snapshot, reasoning effort, prompt, run count, and harness are not specified in the summary table
Overall · Human-aligned Score
Reported by SurveyReview authors / Z.ai · via BenchBoard docs
SurveyReview v1.0 published leaderboard · text survey-paper evaluation · exact candidate snapshot, reasoning effort, prompt, run count, and harness are not specified in the summary table
Overall · Reason Quality Score
Reported by SurveyReview authors / Z.ai · via BenchBoard docs
SurveyReview v1.0 published leaderboard · text survey-paper evaluation · exact candidate snapshot, reasoning effort, prompt, run count, and harness are not specified in the summary table
Readability · MAE
Reported by SurveyReview authors / Z.ai · via BenchBoard docs
SurveyReview v1.0 published leaderboard · text survey-paper evaluation · exact candidate snapshot, reasoning effort, prompt, run count, and harness are not specified in the summary table
Readability · MSE
Reported by SurveyReview authors / Z.ai · via BenchBoard docs
SurveyReview v1.0 published leaderboard · text survey-paper evaluation · exact candidate snapshot, reasoning effort, prompt, run count, and harness are not specified in the summary table
Structure · MSE
Reported by SurveyReview authors / Z.ai · via BenchBoard docs
SurveyReview v1.0 published leaderboard · text survey-paper evaluation · exact candidate snapshot, reasoning effort, prompt, run count, and harness are not specified in the summary table
TAU2Bench2 results
Reported by DeepSeek · via BenchBoard docs · Reported by the model's developer. The source relaying it did not verify it.
Non-thinking mode · temperature 1.0 · 128K context · pass@1
Thinking mode · temperature 1.0 · 128K context · self-play user agent · Airline/Retail/Telecom average
ValsSource: Vals AI
τ²-benchReported by DeepSeek · via BenchBoard docs · Reported by the model's developer. The source relaying it did not verify it.

Data credits: BenchBoard by SYuan03 (github.com/SYuan03/benchboard). MIT covers its code and data organization; no results licence is inferred. Numeric relays retain the original stating-source links. · Data by LLM Stats · OpenEvals/leaderboard-data, MIT (licence) · Epoch AI Benchmarking Hub, CC BY 4.0 (licence) · LMArena, CC BY 4.0 (licence) · BullshitBench by Peter Gostev, MIT (licence) · ChemBench leaderboard, Jablonka group / lamalab-org. Public verified numeric result summaries; results licence not separately stated. · EQ-Bench (Samuel J. Paech) — cite paech2023eqbench · Keelemudelite mõõdupuu: Independent LLM leaderboard for Estonian. Eesti Keele Instituut. https://moodupuu.eki.ee, CC BY 4.0. Latest per-benchmark results selected with partial-run flags retained. (licence) · OpenRouter public benchmark leaderboard and search-lane exports, CC BY 4.0. Fractional source scores converted to percentages; immutable dated snapshots. (licence) · LLM Chess by Maxim Saplin, Apache-2.0 (Copyright [2024] [Maxim Saplin]) (licence) · MineBench, MIT (licence) · Nejumi Leaderboard 4, Weights & Biases Japan; public publisher-selected numeric evaluation results · Open Agent Leaderboard results (Exgentic), CDLA-Permissive-2.0; agreement text at https://cdla.dev/permissive-2-0/. Fractions shown as percentages. (licence) · SnakeBench (Greg Kamradt / ARC Prize Foundation), MIT (licence) · SpreadsheetBench (Renmin University of China and contributors), spreadsheetbench.github.io. · Sierra Research tau2-bench leaderboard submissions (github.com/sierra-research/tau2-bench), MIT (licence) · Vectara Hallucination Leaderboard, Apache-2.0 (licence)

Also known as: alicloud-deepseek-v3.2, azure-deepseek-v3.2, azure_ai/deepseek-v3.2, azure_ai/fw-deepseek-v3.2, bedrock/ap-northeast-1/deepseek.v3.2, bedrock/ap-south-1/deepseek.v3.2, bedrock/ap-southeast-2/deepseek.v3.2, bedrock/ap-southeast-3/deepseek.v3.2, bedrock/eu-central-1/deepseek.v3.2, bedrock/eu-north-1/deepseek.v3.2, bedrock/eu-south-1/deepseek.v3.2, bedrock/eu-west-1/deepseek.v3.2, bedrock/eu-west-2/deepseek.v3.2, bedrock/sa-east-1/deepseek.v3.2, bedrock/us-east-1/deepseek.v3.2, bedrock/us-east-2/deepseek.v3.2, bedrock/us-west-2/deepseek.v3.2, bedrock_mantle/ap-northeast-1/deepseek.v3.2, bedrock_mantle/ap-south-1/deepseek.v3.2, bedrock_mantle/ap-southeast-2/deepseek.v3.2, bedrock_mantle/ap-southeast-3/deepseek.v3.2, bedrock_mantle/deepseek.v3.2, bedrock_mantle/eu-north-1/deepseek.v3.2, bedrock_mantle/eu-west-2/deepseek.v3.2, bedrock_mantle/sa-east-1/deepseek.v3.2, deepinfra/deepseek-ai/deepseek-v3.2, deepinfra/deepseek-v3.2, deepseek-3.2, deepseek-ai/deepseek-v3.2, deepseek-ai/deepseek-v3.2-maas, deepseek-ai/deepseek-v3.2-tee, deepseek-v3-2-251201, deepseek-v3.2-thinking, deepseek-v3.2思考模型, deepseek-v3.2(非思考模式), deepseek-v3p2, deepseek.v3.2, deepseek/deepseek-v3.2, deepseek/deepseek-v3.2-251201, deepseek/deepseek-v3.2-thinking, deepseek/deepseek-v3.2:thinking, deepseek/v3.2, eu.deepseek.v3.2, fireworks_ai/accounts/fireworks/models/deepseek-v3p2, gmi/deepseek-ai/deepseek-v3.2, novita/deepseek-v3.2, novita/deepseek/deepseek-v3.2, pro/deepseek-ai/deepseek-v3.2, sambanova/deepseek-v3.2, siliconflow/deepseek-v3.2, sophnet/deepseek-v3.2, tee/deepseek-v3.2, us.deepseek.v3.2, vertex/deepseek-v3.2, vertex_ai/deepseek-ai/deepseek-v3.2-maas

Like this one. Tap or hover any dotted value, label or icon to see where it comes from and how it is measured. open details when tapped or hovered. How this data works