Maps the full AI model ecosystem by positioning 293 models along score vs price axes. Discover hidden gems that deliver top-tier performance at low cost, and identify overpriced models to avoid.
| # | Model | Score | Cost/1M | Value Ratio |
|---|---|---|---|---|
| 1 | Trinity Miniarcee-ai | 53 | $0.098 | 547.7 |
| 2 | Gemma 3 27BGoogle | 50 | $0.095 | 522.1 |
| 3 | Ministral 3 3B 2512Mistral AI | 49 | $0.100 | 485.0 |
| 4 | Mistral Small 3.2 24BMistral AI | 53 | $0.120 | 440.8 |
| 5 | Nemotron 3 Nano 30B A3BNVIDIA | 48 | $0.125 | 380.8 |
| 6 | Ministral 3 8B 2512Mistral AI | 49 | $0.150 | 328.0 |
| 7 | Seed 1.6 FlashByteDance | 60 | $0.188 | 320.0 |
| 8 | Qwen3 14BAlibaba | 48 | $0.150 | 318.7 |
| 9 | Olmo 3 7B ThinkAllen AI | 49 | $0.160 | 303.8 |
| 10 | Qwen3 32BAlibaba | 48 | $0.160 | 298.8 |
| 11 | Gemini 2.0 Flash LiteGoogle | 55 | $0.188 | 292.8 |
| 12 | GPT-5 NanoOpenAI | 64 | $0.225 | 285.3 |
| 13 | Qwen3 Coder 30B A3B InstructAlibaba | 48 | $0.170 | 282.4 |
| 14 | gpt-oss-safeguard-20bOpenAI | 53 | $0.188 | 282.1 |
| 15 | MiMo-V2-FlashXiaomi | 54 | $0.190 | 282.1 |
| 16 | Qwen3 30B A3BAlibaba | 48 | $0.180 | 265.6 |
| 17 | Llama 4 ScoutMeta | 50 | $0.190 | 264.7 |
| 18 | Gemini 2.5 Flash Lite Preview 09-2025Google | 65 | $0.250 | 261.2 |
| 19 | Qwen3 30B A3B Instruct 2507Alibaba | 50 | $0.195 | 256.4 |
| 20 | Gemini 2.5 Flash LiteGoogle | 64 | $0.250 | 256.0 |
| 21 | Step 3.5 FlashStepFun | 51 | $0.200 | 255.0 |
| 22 | Qwen3.5-FlashAlibaba | 62 | $0.250 | 247.6 |
| 23 | Ministral 3 14B 2512Mistral AI | 49 | $0.200 | 246.0 |
| 24 | Seed-2.0-MiniByteDance | 61 | $0.250 | 244.4 |
| 25 | GPT-4.1 NanoOpenAI | 58 | $0.250 | 230.0 |
| 26 | Gemini 2.0 FlashGoogle | 54 | $0.250 | 217.6 |
| 27 | Qwen3 VL 32B InstructAlibaba | 54 | $0.260 | 208.1 |
| 28 | Tongyi DeepResearch 30B A3BAlibaba | 53 | $0.270 | 197.8 |
| 29 | Qwen3 VL 8B InstructAlibaba | 54 | $0.290 | 186.6 |
| 30 | Grok 4.1 FastxAI | 64 | $0.350 | 182.9 |
| 31 | Grok 4 FastxAI | 64 | $0.350 | 182.9 |
| 32 | Qwen3 VL 30B A3B InstructAlibaba | 54 | $0.325 | 166.5 |
| 33 | DeepSeek V3.2DeepSeek | 53 | $0.325 | 163.7 |
| 34 | DeepSeek V3.2 ExpDeepSeek | 53 | $0.340 | 156.5 |
| 35 | Olmo 3.1 32B ThinkAllen AI | 49 | $0.325 | 149.8 |
| 36 | Olmo 3 32B ThinkAllen AI | 49 | $0.325 | 149.8 |
| 37 | ERNIE 4.5 VL 28B A3BBaidu | 51 | $0.350 | 145.4 |
| 38 | Llama 4 MaverickMeta | 52 | $0.375 | 137.6 |
| 39 | Hunyuan A13B InstructTencent | 48 | $0.355 | 136.1 |
| 40 | Nemotron Nano 12B 2 VLNVIDIA | 50 | $0.400 | 125.8 |
| 41 | Grok 3 MinixAI | 48 | $0.400 | 120.3 |
| 42 | Qwen3 Coder NextAlibaba | 50 | $0.435 | 115.4 |
| 43 | DeepSeek V3.1DeepSeek | 50 | $0.450 | 110.0 |
| 44 | Mercury 2Inception | 53 | $0.500 | 105.6 |
| 45 | Qwen Plus 0728 (thinking)Alibaba | 55 | $0.520 | 105.0 |
| 46 | DeepSeek V3 0324DeepSeek | 49 | $0.485 | 101.9 |
| 47 | LongCat Flash ChatMeituan | 50 | $0.500 | 100.0 |
| 48 | Qwen Plus 0728Alibaba | 51 | $0.520 | 98.3 |
| 49 | KAT-Coder-Pro V1Kuaishou | 51 | $0.517 | 97.8 |
| 50 | Qwen3 VL 235B A22B InstructAlibaba | 49 | $0.540 | 91.5 |
| 51 | Qwen3 Coder FlashAlibaba | 52 | $0.585 | 88.2 |
| 52 | MiniMax M2MiniMax | 54 | $0.627 | 86.5 |
| 53 | Qwen3.5-35B-A3BAlibaba | 61 | $0.731 | 83.1 |
| 54 | Qwen3 VL 8B ThinkingAlibaba | 58 | $0.741 | 78.1 |
| 55 | MiniMax M2.1MiniMax | 48 | $0.610 | 77.9 |
| # | Model | Score | Cost/1M |
|---|---|---|---|
| 1 | GPT-5.4 ProOpenAI | 91 | $105.00 |
| 2 | GPT-5.2 ProOpenAI | 90 | $94.50 |
| 3 | GPT-5 ProOpenAI | 90 | $67.50 |
| 4 | o3 ProOpenAI | 82 | $50.00 |
| 5 | Claude Opus 4.1Anthropic | 81 | $45.00 |
| 6 | o1-proOpenAI | 77 | $375.00 |
| 7 | Claude Opus 4Anthropic | 76 | $45.00 |
| 8 | o3 Deep ResearchOpenAI | 74 | $25.00 |
| 9 | Claude Opus 4.6Anthropic | 71 | $15.00 |
| 10 | Claude Opus 4.5Anthropic | 70 | $15.00 |
| 11 | GPT-5.4OpenAI | 70 | $8.75 |
| 12 | Claude Sonnet 4.5Anthropic | 69 | $9.00 |
| 13 | GPT-5.2OpenAI | 68 | $7.88 |
| 14 | Gemini 3.1 Pro Preview Custom ToolsGoogle | 68 | $7.00 |
| 15 | Gemini 3.1 Pro PreviewGoogle | 68 | $7.00 |
| 16 | Gemini 3 Pro PreviewGoogle | 68 | $7.00 |
| 17 | Claude Sonnet 4.6Anthropic | 68 | $9.00 |
| 18 | GPT-5.1OpenAI | 67 | $5.63 |
| 19 | GPT-5.3-CodexOpenAI | 67 | $7.88 |
| 20 | GPT-5.2-CodexOpenAI | 67 | $7.88 |
| 21 | GPT-5OpenAI | 67 | $5.63 |
| 22 | Gemini 3 Flash PreviewGoogle | 66 | $1.75 |
| 23 | o4 Mini Deep ResearchOpenAI | 66 | $5.00 |
| 24 | GPT-5.1-Codex-MaxOpenAI | 66 | $5.63 |
| 25 | Gemini 3.1 Flash Lite PreviewGoogle | 66 | $0.875 |
| 26 | Gemini 2.5 ProGoogle | 66 | $5.63 |
| 27 | o1OpenAI | 65 | $37.50 |
| 28 | GPT-5 MiniOpenAI | 65 | $1.13 |
| 29 | Gemini 2.5 Pro Preview 05-06Google | 64 | $5.63 |
| 30 | Gemini 2.5 FlashGoogle | 64 | $1.40 |
| 31 | Gemini 2.5 Pro Preview 06-05Google | 64 | $5.63 |
| 32 | Claude Haiku 4.5Anthropic | 63 | $3.00 |
| 33 | GPT-5.3 ChatOpenAI | 62 | $7.88 |
| 34 | Qwen3.5 Plus 2026-02-15Alibaba | 62 | $0.910 |
| 35 | GPT-5.2 ChatOpenAI | 62 | $7.88 |
| 36 | GPT-5.1-CodexOpenAI | 62 | $5.63 |
| 37 | GPT-5 CodexOpenAI | 62 | $5.63 |
| 38 | o3OpenAI | 62 | $5.00 |
| 39 | GPT-5.1 ChatOpenAI | 61 | $5.63 |
| 40 | o4 Mini HighOpenAI | 61 | $2.75 |
| 41 | o4 MiniOpenAI | 61 | $2.75 |
| 42 | Qwen3.5-122B-A10BAlibaba | 61 | $1.17 |
| 43 | Qwen3.5 397B A17BAlibaba | 61 | $1.36 |
| 44 | Claude Sonnet 4Anthropic | 61 | $9.00 |
| 45 | Qwen3.5-27BAlibaba | 61 | $0.878 |
| 46 | Sonar Pro SearchPerplexity | 61 | $9.00 |
| 47 | Nova 2 LiteAmazon | 61 | $1.40 |
| 48 | Seed 1.6ByteDance | 60 | $1.13 |
| 49 | GPT-5.1-Codex-MiniOpenAI | 60 | $1.13 |
| 50 | GPT-4.1OpenAI | 59 | $5.00 |
| 51 | Kimi K2.5Moonshot AI | 59 | $1.33 |
| 52 | Claude 3.7 SonnetAnthropic | 59 | $9.00 |
| 53 | Claude 3.7 Sonnet (thinking)Anthropic | 59 | $9.00 |
| 54 | Grok 4xAI | 58 | $9.00 |
| 55 | GPT-4.1 MiniOpenAI | 58 | $1.000 |
| 56 | GPT-5 ChatOpenAI | 57 | $5.63 |
| 57 | Grok Code Fast 1xAI | 56 | $0.850 |
| 58 | Nova Premier 1.0Amazon | 56 | $7.50 |
| 59 | MiniMax M2.5MiniMax | 54 | $0.747 |
| 60 | Qwen3 Max ThinkingAlibaba | 54 | $2.34 |
| 61 | GPT-4o AudioOpenAI | 53 | $6.25 |
| 62 | Qwen3 Coder PlusAlibaba | 52 | $1.95 |
| 63 | Qwen3 MaxAlibaba | 51 | $3.60 |
| 64 | R1 0528DeepSeek | 51 | $1.30 |
| 65 | GPT AudioOpenAI | 51 | $6.25 |
| 66 | DeepSeek V3.2 SpecialeDeepSeek | 51 | $0.800 |
| 67 | Claude 3.5 SonnetAnthropic | 50 | $18.00 |
| 68 | GPT-4o (2024-11-20)OpenAI | 50 | $6.25 |
| 69 | Mistral Large 3 2512Mistral AI | 50 | $1.00 |
| 70 | MiniMax M1MiniMax | 49 | $1.30 |
| 71 | Qwen3 Coder 480B A35B (exacto)Alibaba | 49 | $1.01 |
| 72 | Jamba Large 1.7AI21 Labs | 49 | $5.00 |
| 73 | Qwen VL MaxAlibaba | 49 | $2.00 |
| 74 | GPT Audio MiniOpenAI | 49 | $1.50 |
| 75 | Sonar ProPerplexity | 49 | $9.00 |
| 76 | Mistral Medium 3.1Mistral AI | 48 | $1.20 |
| 77 | ERNIE 4.5 VL 424B A47B Baidu | 48 | $0.835 |
| 78 | Qwen3 235B A22BAlibaba | 48 | $1.14 |
| 79 | Grok 3xAI | 48 | $9.00 |
| 80 | GPT-4o (extended)OpenAI | 48 | $12.00 |
| # | Model | Score | Cost/1M | Score Deficit |
|---|---|---|---|---|
| 1 | GPT-3.5 Turbo InstructOpenAI | 26 | $1.75 | -21.9 |
| 2 | GPT-3.5 Turbo (older v0613)OpenAI | 29 | $1.50 | -18.3 |
| 3 | Inflection 3 PiInflection | 30 | $6.25 | -17.8 |
| 4 | Inflection 3 ProductivityInflection | 30 | $6.25 | -17.8 |
| 5 | Mistral LargeMistral AI | 30 | $4.00 | -17.8 |
| 6 | Llama 3.1 405B (base)Meta | 30 | $4.00 | -17.3 |
| 7 | Mixtral 8x22B InstructMistral AI | 30 | $4.00 | -17.3 |
| 8 | GPT-3.5 TurboOpenAI | 31 | $1.00 | -17.0 |
| 9 | GPT-3.5 Turbo 16kOpenAI | 31 | $3.50 | -16.4 |
| 10 | Llemma 7beleutherai | 33 | $1.00 | -14.6 |
| 11 | Llama 3.1 405B InstructMeta | 33 | $4.00 | -14.5 |
| 12 | Claude 3 HaikuAnthropic | 35 | $0.750 | -12.6 |
| 13 | Mistral Large 2411Mistral AI | 37 | $4.00 | -10.7 |
| 14 | Mistral Large 2407Mistral AI | 37 | $4.00 | -10.7 |
| 15 | Command R+ (08-2024)Cohere | 37 | $6.25 | -10.3 |
| 16 | Maestro Reasoningarcee-ai | 39 | $2.10 | -8.5 |
| 17 | SonarPerplexity | 39 | $1.00 | -8.4 |
| 18 | MiniMax M2-herMiniMax | 39 | $0.750 | -8.3 |
| 19 | Llama 3.1 Nemotron 70B InstructNVIDIA | 40 | $1.20 | -7.9 |
| 20 | Aion-1.0-Miniaion-labs | 40 | $1.05 | -7.8 |
| 21 | GPT-4 Turbo PreviewOpenAI | 40 | $20.00 | -7.8 |
| 22 | GPT-4 Turbo (older v1106)OpenAI | 40 | $20.00 | -7.8 |
| 23 | Sonar Deep ResearchPerplexity | 40 | $5.00 | -7.3 |
| 24 | Aion-1.0aion-labs | 41 | $6.00 | -6.2 |
| 25 | Qwen-Max Alibaba | 41 | $2.60 | -6.1 |
| 26 | R1DeepSeek | 42 | $1.60 | -5.6 |
| 27 | Pixtral Large 2411Mistral AI | 42 | $4.00 | -5.5 |
| 28 | Devstral MediumMistral AI | 42 | $1.20 | -5.4 |
| 29 | Kimi K2 0711Moonshot AI | 42 | $1.38 | -5.3 |
| 30 | R1 Distill Llama 70BDeepSeek | 42 | $0.750 | -5.2 |
| 31 | Command ACohere | 43 | $6.25 | -5.0 |
| 32 | Virtuoso Largearcee-ai | 43 | $0.975 | -4.9 |
| 33 | Nova Pro 1.0Amazon | 43 | $2.00 | -4.3 |
| 34 | Qwen2.5 VL 72B InstructAlibaba | 43 | $0.800 | -4.2 |
| 35 | Cogito v2.1 671Bdeepcogito | 44 | $1.25 | -4.0 |
| 36 | Kimi K2 0905Moonshot AI | 44 | $1.20 | -3.8 |
| 37 | Palmyra X5Writer | 44 | $3.30 | -3.2 |
| 38 | Devstral 2 2512Mistral AI | 44 | $1.20 | -3.1 |
| 39 | GPT-4 (older v0314)OpenAI | 44 | $45.00 | -3.1 |
| 40 | GPT-4OpenAI | 44 | $45.00 | -3.1 |
| 41 | Kimi K2 0905 (exacto)Moonshot AI | 45 | $1.55 | -3.0 |
| 42 | GPT-4oOpenAI | 45 | $6.25 | -3.0 |
| 43 | GPT-4o (2024-05-13)OpenAI | 45 | $10.00 | -2.8 |
| 44 | Qwen-PlusAlibaba | 45 | $0.800 | -2.3 |
| 45 | Sonar Reasoning ProPerplexity | 45 | $5.00 | -2.1 |
| 46 | Aion-2.0aion-labs | 46 | $1.20 | -1.9 |
| 47 | Mistral Medium 3Mistral AI | 46 | $1.20 | -1.9 |
| 48 | GPT-4o Search PreviewOpenAI | 46 | $6.25 | -1.6 |
| 49 | GPT-4 TurboOpenAI | 46 | $20.00 | -1.6 |
| 50 | Claude 3.5 HaikuAnthropic | 46 | $2.40 | -1.4 |
| 51 | Grok 3 BetaxAI | 47 | $9.00 | -1.0 |
| 52 | o3 MiniOpenAI | 47 | $2.75 | -0.7 |
| 53 | GPT-4o (2024-08-06)OpenAI | 47 | $6.25 | -0.7 |
| 54 | o3 Mini HighOpenAI | 47 | $2.75 | -0.4 |
| 55 | Kimi K2 ThinkingMoonshot AI | 47 | $1.23 | -0.2 |
| # | Model | Score |
|---|---|---|
| 1 | Qwen3 VL 30B A3B ThinkingAlibaba | 69 |
| 2 | Qwen3 VL 235B A22B ThinkingAlibaba | 69 |
| 3 | Nemotron Nano 12B 2 VL (free)NVIDIA | 64 |
| 4 | Step 3.5 Flash (free)StepFun | 58 |
| 5 | Qwen3 235B A22B Thinking 2507Alibaba | 57 |
| 6 | gpt-oss-120b (free)OpenAI | 56 |
| 7 | gpt-oss-20b (free)OpenAI | 56 |
| 8 | Gemma 3 27B (free)Google | 56 |
| 9 | Trinity Large Preview (free)arcee-ai | 54 |
| 10 | Trinity Mini (free)arcee-ai | 54 |
| 11 | Nemotron Nano 9B V2 (free)NVIDIA | 54 |
| 12 | Qwen3 Coder 480B A35B (free)Alibaba | 54 |
| 13 | Nemotron 3 Nano 30B A3B (free)NVIDIA | 51 |
| 14 | Qwen3 Next 80B A3B Instruct (free)Alibaba | 51 |
| 15 | Mistral Small 3.1 24B (free)Mistral AI | 51 |
| 16 | Gemma 3 4B (free)Google | 51 |
| 17 | Qwen3 4B (free)Alibaba | 50 |
| 18 | Gemma 3 12B (free)Google | 47 |
| 19 | Gemma 3n 2B (free)Google | 46 |
| 20 | Llama 3.3 70B Instruct (free)Meta | 46 |
| 21 | LFM2.5-1.2B-Thinking (free)Liquid AI | 46 |
| 22 | Gemma 3n 4B (free)Google | 45 |
| 23 | LFM2.5-1.2B-Instruct (free)Liquid AI | 42 |
| 24 | Llama 3.2 3B Instruct (free)Meta | 34 |
The median score is 48 and the median cost is $0.747/1M tokens. These thresholds define the quadrant boundaries.
20.4% of paid models are Hidden Gems -- great quality at prices below the median. These are the models worth paying attention to.
55 models sit in the Overpriced quadrant -- below-median performance at above-median prices. Consider alternatives before choosing these.
The best free model (Qwen3 VL 30B A3B Thinking) scores 69, outperforming 257 paid models.