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.
Gemma 4 31BXing4.0-29B-A4BNemotron 3 Ultra
Focus

Writing & chat: quality vs price

Compared models are ringed; the other 143 ranked here are greyed.

Lower price is better. Pareto frontier: Gemma 4 E4B, Qwen3.5 4B, Gemma 4 12B, Qwen3.5 9B, Gemma 4 26B A4B, GPT-6 Luna, Qwen3.8 Flash-Next, MiMo-V2.6-Pro, Step 5 Preview, GLM-5.3, Kimi K3, Claude Opus 5.5, Claude Opus 5, Claude Fable 5.1. 52 models are not plotted: Intern-S2-Preview (35B-A3B), Xing4.0-29B-A4B, Nex-N2-Pro, Nemotron Cascade 2 30B A3B, Apriel-v1.6-15B-Thinker, Command A+, EXAONE 4.5 33B, HyperNova 60B 2605, DiffusionGemma 26B A4B, K2 Think V2, K2-V2, Solar Open 100B, Step3 VL 10B, Tri-21B-Think, Falcon-H1R-7B, North Mini Code, NVIDIA Nemotron 3 Nano 4B, LFM2.5-8B-A1B, LongCat Flash Lite, MiniCPM5-1B, Jamba Reasoning 3B, LFM2 24B A2B, Nanbeige4.1-3B, HyperCLOVA X SEED Think (32B), INTELLECT-3, Olmo 3 7B Think, LFM2.5-1.2B-Instruct, Llama 3.1 Nemotron Ultra 253B v1, LFM2.5-1.2B-Thinking, Devstral 2, Olmo 3.1 32B Instruct, Olmo 3.1 32B Think, Gemma 4 E2B, Qwen3.5 2B, EXAONE 4.0 32B, Devstral Small 2, Jamba 1.7 Large, MiniCPM-V 4.6 1.3B, LFM2.5-VL-1.6B, Hermes 4 - Llama-3.1 70B, Molmo2-8B, Jamba 1.7 Mini, Kimi Linear 48B A3B Instruct, Exaone 4.0 1.2B, Granite 4.0 H 1B, Granite 4.0 Micro, Qwen3.5 0.8B, Molmo 7B-D, Tiny Aya Global, Granite 4.0 H 350M, Granite 4.0 350M, Gemma 3 270M.

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

Intern-S2-Preview (35B-A3B), Xing4.0-29B-A4B, Nex-N2-Pro, Nemotron Cascade 2 30B A3B, Apriel-v1.6-15B-Thinker, Command A+, EXAONE 4.5 33B, HyperNova 60B 2605, DiffusionGemma 26B A4B, K2 Think V2, K2-V2, Solar Open 100B, Step3 VL 10B, Tri-21B-Think, Falcon-H1R-7B, North Mini Code, NVIDIA Nemotron 3 Nano 4B, LFM2.5-8B-A1B, LongCat Flash Lite, MiniCPM5-1B, Jamba Reasoning 3B, LFM2 24B A2B, Nanbeige4.1-3B, HyperCLOVA X SEED Think (32B), INTELLECT-3, Olmo 3 7B Think, LFM2.5-1.2B-Instruct, Llama 3.1 Nemotron Ultra 253B v1, LFM2.5-1.2B-Thinking, Devstral 2, Olmo 3.1 32B Instruct, Olmo 3.1 32B Think, Gemma 4 E2B, Qwen3.5 2B, EXAONE 4.0 32B, Devstral Small 2, Jamba 1.7 Large, MiniCPM-V 4.6 1.3B, LFM2.5-VL-1.6B, Hermes 4 - Llama-3.1 70B, Molmo2-8B, Jamba 1.7 Mini, Kimi Linear 48B A3B Instruct, Exaone 4.0 1.2B, Granite 4.0 H 1B, Granite 4.0 Micro, Qwen3.5 0.8B, Molmo 7B-D, Tiny Aya Global, Granite 4.0 H 350M, Granite 4.0 350M, Gemma 3 270M

  • Best-value frontier (nothing is both cheaper and better)
  • Evidencestrong → weak
  • Estimated from other categories
  • Gemma 4 31B
    Quality
    63.5
    Rank
    #39/146
    Price
    $0.21/M tok
    Speed
    35tok/s
    In$0.14▾Out$0.40▾/M tok

    vs Xing4.0-29B-A4B: −3.9 quality · price n/a

  • Xing4.0-29B-A4B
    Quality
    67.4
    Rank
    #32/146
    Price
    —
    Speed
    —

    vs Gemma 4 31B: +3.9 quality · price n/a

  • Nemotron 3 Ultra
    Quality
    60.7
    Rank
    #46/146
    Price
    $1.07/M tok
    Speed
    215tok/s
    In$0.60Out$2.50/M tok

    vs Xing4.0-29B-A4B: −6.7 quality · price n/a

Overview

Lab
Google DeepMind
China Telecom AI (TeleAI)
NVIDIA
Released
Apr 2, 2026
Sep 17, 2026
Jun 4, 2026
Weights
Open
Open
Open
Input price
$/M tok
$0.14Best
—
$0.60
Output price
$/M tok
$0.40Best
—
$2.50
Blended price
$/M tok · 3:1 in:out
$0.205Best
—
$1.075
Output speed
tok/s
35
⚡ SambaNova 200
—
215Best
Context
tokens
256K
262KBest
262KBest

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.

39.6±6.5
#44/103Mixed
62.1±10.6Best
#19/103Lab-reported
37.1±5.4
#51/103Mixed
63.5±3.6
#39/146Mixed
67.4±9.7Best
#32/146Lab-reported
60.7±3.3
#46/146Verified
57.5±8.2
#34/82Verified
Not ranked
Not ranked
59.6±4.5
#47/182Mixed
59.8±10.0Best
#46/182Lab-reported
58.9±3.7
#48/182Verified
30.4±4.2
#88/173Verified
Not ranked
31.3±3.4Best
#86/173Mixed

Writing & chat benchmarks

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

Benchmarks in other categories (4)

Coding benchmarks

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

~last/15below measured range (<41.1)Estimated
estimated from related benchmarks
~last/15below measured range (<41.1)Estimated
estimated from related benchmarks

Vision benchmarks

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

Hard reasoning benchmarks

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

~last/5below measured range (<47.5%)Estimated
estimated from related benchmarks
—
~last/5below measured range (<47.5%)Estimated
estimated from related benchmarks