Skip to content
Model Pareto

Compare models

Side by side: quality per category, price, speed, and every benchmark cell with where it came from. Pick up to four.
Apertus 70B InstructOlmo 3 7B InstructJamba 1.7 Mini
Focus

Hard reasoning: quality vs price

Compared models are ringed; the other 179 ranked here are greyed.

Lower price is better. Pareto frontier: Gemma 4 E4B, Sarvam 30B, Qwen3.5 4B, Sarvam 105B, Ling 3.0 Flash, MiMo-V2.5, GPT-6 Luna, Qwen3.8 Flash-Next, GLM-5.3 Flash, MiMo-V2.6-Pro, Muse Spark 1.3, Claude Sonnet 5.5, Claude Opus 5.5. 70 models are not plotted: Motif 3, Nex-N2-Pro, K2 Horizon 375B A23B, A.X-K2, Solar Open2 250B, Xing4.0-29B-A4B, K2 Horizon MoVA 36B A4B, Intern-S2-Preview (35B-A3B), K-EXAONE 2.0 0803, G9v3-39A5B, K2 Horizon 7B, Phi-4-reasoning-plus, EXAONE 4.5 33B, Nanbeige4.1-3B, HyperNova 60B 2605, Nemotron Cascade 2 30B A3B, INTELLECT-3, Apriel-v1.6-15B-Thinker, Command A+, K2 Horizon 3.7B, Ling 3.0 Tiny, K2 Think V2, Step3 VL 10B, DiffusionGemma 26B A4B, K2-V2, North Mini Code, Falcon-H1R-7B, Solar Open 100B, Llama 3.1 Nemotron Ultra 253B v1, MiniCPM5-2B, LongCat Flash Lite, HyperCLOVA X SEED Think (32B), Tri-21B-Think, EXAONE 4.0 32B, Olmo 3.1 32B Think, LFM2.5-2.6B, Devstral 2, LFM2.5-8B-A1B, Olmo 3.1 32B Instruct, Olmo 3 7B Think, NVIDIA Nemotron 3 Nano 4B, Devstral Small 2, Phi-4-mini-flash-reasoning, Falcon-H1-34B-Instruct, Qwen3.5 2B, Hermes 4 - Llama-3.1 70B, LFM2 24B A2B, Exaone 4.0 1.2B, Llama 3.2 Instruct 90B (Vision), Gemma 4 E2B, Molmo2-8B, LFM2.5-1.2B-Thinking, LFM2.5-1.2B-Instruct, MiniCPM5-1B, Jamba 1.7 Large, MiniCPM-V 4.6 1.3B, Kimi Linear 48B A3B Instruct, Granite 4.0 Micro, Jamba Reasoning 3B, Jamba 1.7 Mini, Phi-4 Multimodal Instruct, Tiny Aya Global, LFM2.5-VL-1.6B, Granite 4.0 H 350M, K2 Horizon 0.9B, Qwen3.5 0.8B, Granite 4.0 350M, Granite 4.0 H 1B, Molmo 7B-D, Gemma 3 270M.

70 models with no price data — shown in the strip at the left edge

Motif 3, Nex-N2-Pro, K2 Horizon 375B A23B, A.X-K2, Solar Open2 250B, Xing4.0-29B-A4B, K2 Horizon MoVA 36B A4B, Intern-S2-Preview (35B-A3B), K-EXAONE 2.0 0803, G9v3-39A5B, K2 Horizon 7B, Phi-4-reasoning-plus, EXAONE 4.5 33B, Nanbeige4.1-3B, HyperNova 60B 2605, Nemotron Cascade 2 30B A3B, INTELLECT-3, Apriel-v1.6-15B-Thinker, Command A+, K2 Horizon 3.7B, Ling 3.0 Tiny, K2 Think V2, Step3 VL 10B, DiffusionGemma 26B A4B, K2-V2, North Mini Code, Falcon-H1R-7B, Solar Open 100B, Llama 3.1 Nemotron Ultra 253B v1, MiniCPM5-2B, LongCat Flash Lite, HyperCLOVA X SEED Think (32B), Tri-21B-Think, EXAONE 4.0 32B, Olmo 3.1 32B Think, LFM2.5-2.6B, Devstral 2, LFM2.5-8B-A1B, Olmo 3.1 32B Instruct, Olmo 3 7B Think, NVIDIA Nemotron 3 Nano 4B, Devstral Small 2, Phi-4-mini-flash-reasoning, Falcon-H1-34B-Instruct, Qwen3.5 2B, Hermes 4 - Llama-3.1 70B, LFM2 24B A2B, Exaone 4.0 1.2B, Llama 3.2 Instruct 90B (Vision), Gemma 4 E2B, Molmo2-8B, LFM2.5-1.2B-Thinking, LFM2.5-1.2B-Instruct, MiniCPM5-1B, Jamba 1.7 Large, MiniCPM-V 4.6 1.3B, Kimi Linear 48B A3B Instruct, Granite 4.0 Micro, Jamba Reasoning 3B, Jamba 1.7 Mini, Phi-4 Multimodal Instruct, Tiny Aya Global, LFM2.5-VL-1.6B, Granite 4.0 H 350M, K2 Horizon 0.9B, Qwen3.5 0.8B, Granite 4.0 350M, Granite 4.0 H 1B, Molmo 7B-D, Gemma 3 270M

  • Best-value frontier (nothing is both cheaper and better)
  • Evidencestrong → weak
  • Estimated from other categories
  • Apertus 70B Instruct
    Quality
    8.5
    Rank
    #174/182
    Price
    $1.34/M tok
    Speed
    —
    In$0.82Out$2.92/M tok

    vs Olmo 3 7B Instruct: −5.3 quality · 11× price

  • Olmo 3 7B Instruct
    Quality
    13.8
    Rank
    #154/182
    Price
    $0.13/M tok
    Speed
    —
    In$0.10▾Out$0.20▾/M tok

    vs Jamba 1.7 Mini: +4.7 quality · price n/a

  • Jamba 1.7 Mini
    Quality
    9.1
    Rank
    #169/182
    Price
    —
    Speed
    —

    vs Olmo 3 7B Instruct: −4.7 quality · price n/a

Overview

Metric
Apertus 70B Instruct
Olmo 3 7B Instruct
Jamba 1.7 Mini
Lab
Swiss AI Initiative (Apertus)
Ai2 (Allen Institute)
AI21 Labs
Released
Sep 2, 2025
Nov 20, 2025
Jul 7, 2025
Weights
Open
Open
Open
Input price
$/M tok
$0.82
$0.10Best
—
Output price
$/M tok
$2.92
$0.20Best
—
Blended price
$/M tok · 3:1 in:out
$1.345
$0.125Best
—
Context
tokens
66K
66K
258KBest

Quality by category

0–100 within each category (100 = best tracked model). Rank is among all models ranked there. Click a row to focus it.

Metric
Apertus 70B Instruct
Olmo 3 7B Instruct
Jamba 1.7 Mini
Writing & chat
14.3±9.5
#137/146Verified
20.0±9.5Best
#121/146Verified
17.9±9.5
#130/146Verified
Hard reasoning
8.5±5.8
#174/182Verified
13.8±5.8Best
#154/182Verified
9.1±5.8
#169/182Verified
Agents
16.5±10.0
#148/173Verified
16.5±10.0Best
#147/173Verified
16.4±10.0
#149/173Verified

Hard reasoning benchmarks

Rank among models with a published score, and the leaderboard around each model. ~ italic = no published score, estimated from related benchmarks.

Metric
Apertus 70B Instruct
Olmo 3 7B Instruct
Jamba 1.7 Mini
GPQA Diamond
day 0
#164/17227.2%Verified
  1. #1GPT-6 Astra96.1%
  2. #163MiniCPM5-1B27.8%
  3. #164Apertus 70B Instruct27.2%
  4. #165Granite 4.0 H 1B26.3%
artificialanalysis.ai · 2026-09-24
#149/17240.0%BestVerified
  1. #1GPT-6 Astra96.1%
  2. #148Gemma 4 E2B40.5%
  3. #149Olmo 3 7B Instruct40.0%
  4. #150Jamba 1.7 Large39.0%
artificialanalysis.ai · 2026-09-24
#157/17232.2%Verified
  1. #1GPT-6 Astra96.1%
  2. #156LFM2.5-1.2B-Instruct32.6%
  3. #157Jamba 1.7 Mini32.2%
  4. #158Phi-4 Multimodal Instruct31.5%
artificialanalysis.ai · 2026-09-24
Humanity's Last Exam
day 0
#130/1785.5%Verified
  1. #1Claude Opus 5.561.4%
  2. #129Exaone 4.0 1.2B5.7%
  3. #130Apertus 70B Instruct5.5%
  4. #130Granite 4.0 350M5.5%
artificialanalysis.ai · 2026-09-24
#127/1785.8%BestVerified
  1. #1Claude Opus 5.561.4%
  2. #127LongCat Flash Lite5.8%
  3. #127Olmo 3 7B Instruct5.8%
  4. #129Exaone 4.0 1.2B5.7%
artificialanalysis.ai · 2026-09-24
#159/1784.4%Verified
  1. #1Claude Opus 5.561.4%
  2. #158Llama 3.2 Instruct 90B (Vision)4.5%
  3. #159Jamba 1.7 Mini4.4%
  4. #159Reka Flash 34.4%
artificialanalysis.ai · 2026-09-24
AIME (latest)
day 0
~#21/25~19.4%Estimated
  1. #1Inkling97.1%
  2. #20Ministral 3 3B22.0%
  3. ~#21Apertus 70B InstructEstimated19.4%
  4. #21Llama 4 Maverick19.3%
estimated from related benchmarks
~#20/25~25.0%Estimated
  1. #1Inkling97.1%
  2. #19Ministral 3 14B30.0%
  3. ~#20Olmo 3 7B InstructEstimated25.0%
  4. #20Ministral 3 3B22.0%
estimated from related benchmarks
~#22/25~18.4%Estimated
  1. #1Inkling97.1%
  2. #21Llama 4 Maverick19.3%
  3. ~#22Jamba 1.7 MiniEstimated18.4%
  4. #22Falcon-H1-34B-Instruct16.7%
estimated from related benchmarks
AA Intelligence Index
third-party
~#74/75~4.5Estimated
  1. #1Claude Opus 5.557.6
  2. #73Ministral 3 3B4.8
  3. ~#74Apertus 70B InstructEstimated4.5
  4. #74K2 Horizon 0.9B3.0
estimated from related benchmarks
~#73/75~5.0Estimated
  1. #1Claude Opus 5.557.6
  2. #72Ministral 3 8B5.5
  3. ~#73Olmo 3 7B InstructEstimated5.0
  4. #73Ministral 3 3B4.8
estimated from related benchmarks
~#74/75~4.2Estimated
  1. #1Claude Opus 5.557.6
  2. #73Ministral 3 3B4.8
  3. ~#74Jamba 1.7 MiniEstimated4.2
  4. #74K2 Horizon 0.9B3.0
estimated from related benchmarks
Benchmarks in other categories (2)

Writing & chat benchmarks

Rank among models with a published score, and the leaderboard around each model. ~ italic = no published score, estimated from related benchmarks.

Metric
Apertus 70B Instruct
Olmo 3 7B Instruct
Jamba 1.7 Mini
IFBench
day 0
#115/12525.9%Verified
  1. #1MiniMax-M382.9%
  2. #114Granite 4.0 H 1B26.2%
  3. #115Apertus 70B Instruct25.9%
  4. #116Exaone 4.0 1.2B25.3%
artificialanalysis.ai · 2026-09-24
#93/12532.8%BestVerified
  1. #1MiniMax-M382.9%
  2. #92LFM2.5-VL-1.6B33.1%
  3. #93Olmo 3 7B Instruct32.8%
  4. #94Hermes 4 - Llama-3.1 405B32.7%
artificialanalysis.ai · 2026-09-24
#98/12531.4%Verified
  1. #1MiniMax-M382.9%
  2. #97Granite 4.0 H Small31.5%
  3. #98Jamba 1.7 Mini31.4%
  4. #99Devstral Small 231.2%
artificialanalysis.ai · 2026-09-24
MMMLU
day 0
~last/5below measured range (<81.3%)Estimated
  1. #1Gemini 3.1 Pro92.6%
  2. #4gpt-oss-120b81.3%
  3. ~#5Apertus 70B InstructEstimated<81.3%
estimated from related benchmarks
~last/5below measured range (<81.3%)Estimated
  1. #1Gemini 3.1 Pro92.6%
  2. #4gpt-oss-120b81.3%
  3. ~#5Olmo 3 7B InstructEstimated<81.3%
estimated from related benchmarks
~last/5below measured range (<81.3%)Estimated
  1. #1Gemini 3.1 Pro92.6%
  2. #4gpt-oss-120b81.3%
  3. ~#5Jamba 1.7 MiniEstimated<81.3%
estimated from related benchmarks
LMArena Text, Non-English (Elo)
third-party
~#52/54~1273Estimated
  1. #1Claude Opus 5.51501
  2. #51NVIDIA Nemotron 3 Nano 30B A3B1286
  3. ~#52Apertus 70B InstructEstimated1273
  4. #52Phi-41233
estimated from related benchmarks
~#49/54~1295Estimated
  1. #1Claude Opus 5.51501
  2. #48gpt-oss-20b1299
  3. ~#49Olmo 3 7B InstructEstimated1295
  4. #49Olmo 3.1 32B Instruct1294
estimated from related benchmarks
~#52/54~1284Estimated
  1. #1Claude Opus 5.51501
  2. #51NVIDIA Nemotron 3 Nano 30B A3B1286
  3. ~#52Jamba 1.7 MiniEstimated1284
  4. #52Phi-41233
estimated from related benchmarks
LMArena Text (Elo)
third-party
~#52/54~1302Estimated
  1. #1Claude Opus 5.51509
  2. #51NVIDIA Nemotron 3 Nano 30B A3B1314
  3. ~#52Apertus 70B InstructEstimated1302
  4. #52Olmo 3.1 32B Think1286
estimated from related benchmarks
~#49/54~1321Estimated
  1. #1Claude Opus 5.51509
  2. #48Molmo2-8BFew votes1322
  3. ~#49Olmo 3 7B InstructEstimated1321
  4. #49Ring-flash-2.01321
estimated from related benchmarks
~#52/54~1312Estimated
  1. #1Claude Opus 5.51509
  2. #51NVIDIA Nemotron 3 Nano 30B A3B1314
  3. ~#52Jamba 1.7 MiniEstimated1312
  4. #52Olmo 3.1 32B Think1286
estimated from related benchmarks
EQ-Bench Creative Writing v3 (Elo)
third-party
~#23/26~1411Estimated
  1. #1GPT-6 Astra2173
  2. #22Mistral Large 31412
  3. ~#23Apertus 70B InstructEstimated1411
  4. #23Gemma 4 31B1368
estimated from related benchmarks
~#22/26~1428Estimated
  1. #1GPT-6 Astra2173
  2. #21Gemini 3.1 Pro1491
  3. ~#22Olmo 3 7B InstructEstimated1428
  4. #22Mistral Large 31412
estimated from related benchmarks
~#22/26~1425Estimated
  1. #1GPT-6 Astra2173
  2. #21Gemini 3.1 Pro1491
  3. ~#22Jamba 1.7 MiniEstimated1425
  4. #22Mistral Large 31412
estimated from related benchmarks

Agents benchmarks

Rank among models with a published score, and the leaderboard around each model. ~ italic = no published score, estimated from related benchmarks.

Metric
Apertus 70B Instruct
Olmo 3 7B Instruct
Jamba 1.7 Mini
τ²-bench (Telecom)
day 0
#95/11512.9%BestVerified
  1. #1Step 3.7 Flash98.5%
  2. #94Granite 4.0 350M13.2%
  3. #95Apertus 70B Instruct12.9%
  4. #96Granite 4.0 Micro12.6%
artificialanalysis.ai · 2026-09-24
#96/11512.6%Verified
  1. #1Step 3.7 Flash98.5%
  2. #96Jamba 1.7 Mini12.6%
  3. #96Olmo 3 7B Instruct12.6%
  4. #99Apertus 8B Instruct11.4%
artificialanalysis.ai · 2026-09-24
#96/11512.6%Verified
  1. #1Step 3.7 Flash98.5%
  2. #96Granite 4.0 Micro12.6%
  3. #96Jamba 1.7 Mini12.6%
  4. #96Olmo 3 7B Instruct12.6%
artificialanalysis.ai · 2026-09-24
OSWorld-Verified
day 0
~#11/12~62.1%Estimated
  1. #1Kimi K384.8%
  2. #10Muse Glimmer65.9%
  3. ~#11Apertus 70B InstructEstimated62.1%
  4. #11Claude Haiku 4.550.7%
estimated from related benchmarks
~#11/12~62.3%Estimated
  1. #1Kimi K384.8%
  2. #10Muse Glimmer65.9%
  3. ~#11Olmo 3 7B InstructEstimated62.3%
  4. #11Claude Haiku 4.550.7%
estimated from related benchmarks
~#11/12~62.2%Estimated
  1. #1Kimi K384.8%
  2. #10Muse Glimmer65.9%
  3. ~#11Jamba 1.7 MiniEstimated62.2%
  4. #11Claude Haiku 4.550.7%
estimated from related benchmarks
OSWorld 2.0
day 0
~#3/5~53.8%Estimated
  1. #1GLM-5.3 Flash59.1%
  2. #2Kimi K358.3%
  3. ~#3Apertus 70B InstructEstimated53.8%
  4. #3GPT-5.6 Terra50.2%
estimated from related benchmarks
~#3/5~53.8%Estimated
  1. #1GLM-5.3 Flash59.1%
  2. #2Kimi K358.3%
  3. ~#3Olmo 3 7B InstructEstimated53.8%
  4. #3GPT-5.6 Terra50.2%
estimated from related benchmarks
~#3/5~53.8%Estimated
  1. #1GLM-5.3 Flash59.1%
  2. #2Kimi K358.3%
  3. ~#3Jamba 1.7 MiniEstimated53.8%
  4. #3GPT-5.6 Terra50.2%
estimated from related benchmarks
BrowseComp
day 0
~#17/18~31.0%Estimated
  1. #1Atria Dawn Preview92.5%
  2. #16Nemotron 3 Ultra44.4%
  3. ~#17Apertus 70B InstructEstimated31.0%
  4. #17Mistral Small 421.3%
estimated from related benchmarks
~#17/18~31.1%Estimated
  1. #1Atria Dawn Preview92.5%
  2. #16Nemotron 3 Ultra44.4%
  3. ~#17Olmo 3 7B InstructEstimated31.1%
  4. #17Mistral Small 421.3%
estimated from related benchmarks
~#17/18~31.1%Estimated
  1. #1Atria Dawn Preview92.5%
  2. #16Nemotron 3 Ultra44.4%
  3. ~#17Jamba 1.7 MiniEstimated31.1%
  4. #17Mistral Small 421.3%
estimated from related benchmarks
GDPval-AA (Elo)
third-party
~#56/87~690Estimated
  1. #1Claude Opus 5.51846
  2. #55MiniCPM5-2B694
  3. ~#56Apertus 70B InstructEstimated690
  4. #56Nemotron 3.5 Lightning624
estimated from related benchmarks
~#56/87~688Estimated
  1. #1Claude Opus 5.51846
  2. #55MiniCPM5-2B694
  3. ~#56Olmo 3 7B InstructEstimated688
  4. #56Nemotron 3.5 Lightning624
estimated from related benchmarks
~#56/87~688Estimated
  1. #1Claude Opus 5.51846
  2. #55MiniCPM5-2B694
  3. ~#56Jamba 1.7 MiniEstimated688
  4. #56Nemotron 3.5 Lightning624
estimated from related benchmarks
AutomationBench-AA
third-party
~#74/80~0.3%Estimated
  1. #1Claude Opus 5.569.5%
  2. #72Granite 4.2 3B0.4%
  3. ~#74Apertus 70B InstructEstimated0.3%
  4. #74gpt-oss-120b0.2%
estimated from related benchmarks
~#74/80~0.3%Estimated
  1. #1Claude Opus 5.569.5%
  2. #72Granite 4.2 3B0.4%
  3. ~#74Olmo 3 7B InstructEstimated0.3%
  4. #74gpt-oss-120b0.2%
estimated from related benchmarks
~#74/80~0.3%Estimated
  1. #1Claude Opus 5.569.5%
  2. #72Granite 4.2 3B0.4%
  3. ~#74Jamba 1.7 MiniEstimated0.3%
  4. #74gpt-oss-120b0.2%
estimated from related benchmarks
τ-Bench Banking (AA)
third-party
~last/89below measured range (<3.3%)Estimated
  1. #1Muse Spark 1.350.5%
  2. #88Llama 4 Scout3.3%
  3. ~#89Apertus 70B InstructEstimated<3.3%
estimated from related benchmarks
~last/89below measured range (<3.3%)Estimated
  1. #1Muse Spark 1.350.5%
  2. #88Llama 4 Scout3.3%
  3. ~#89Olmo 3 7B InstructEstimated<3.3%
estimated from related benchmarks
~last/89below measured range (<3.3%)Estimated
  1. #1Muse Spark 1.350.5%
  2. #88Llama 4 Scout3.3%
  3. ~#89Jamba 1.7 MiniEstimated<3.3%
estimated from related benchmarks