AI Agent Store/Model comparisonsBuild an AI worker ↗THE MODEL SELECTION SERIES
An agent model needs to interpret a goal, obey constraints, use tools and recover from mistakes. Our General agent profile weights reasoning, instruction following, agentic coding and data work to build a practical first shortlist.
Data reviewed · LiveBench question set: 25 June 2026 · View every source ↗ · Download the data ↓
Use the General agent profile for mixed workflows. Switch to Coding assistant for repository work, Research & analysis for structured information, or Writing & support for language-heavy tasks. These editorial weights express priorities; they do not reproduce your production environment.
A tool-calling API can return a correctly shaped function call while choosing the wrong action. Check the catalog support flag, read Berkeley’s separate tool-use results, and test the exact tools, schemas and permissions your own agent will use.
A cheap token rate can still produce an expensive agent if it needs repeated attempts or very long reasoning. Enter total billed output, including hidden reasoning where charged. Track success rate, retries and human review alongside the bill.
Take 20–50 real tasks, remove private data where appropriate, and define what a successful result looks like before running models. Keep the prompt, tool definitions, reasoning budget and retry policy fixed. Include ambiguous requests, missing inputs and failing tools.
YOUR WORKLOAD. YOUR PRIORITIES.
Tokens are the small pieces of text a model reads or produces. Include charged thinking tokens in output.
This estimate uses your token counts and standard text API rates. It excludes cache discounts, tool fees, retries and provider premiums. If your input and output exceed a model’s listed capacity, we leave its cost and Value rating blank. A large capacity does not guarantee that it will use every detail correctly.
Shaded zone: 75.2+ points and $21 or less. Numbers mark models with strong price–quality tradeoffs in or near this zone. On the dotted line, no other shown model is cheaper with an equal or higher score, or scores higher at the same price. Your chosen measure, usage and filters change this shortlist.
Practical Value · General agent
70% Agent Fit + 30% affordability. Rankings use your workload. Filtering the list does not change how scores are calculated.
THE EVIDENCE, SIDE BY SIDE
Anthropic · max
Anthropic · max
Meta · xhigh
Anthropic · source ID max; display xhigh
DeepSeek · max
OpenAI · max
Meta · max · version 1.0
OpenAI · max
Moonshot AI · source default
OpenAI · source ID max; display xhigh
Alibaba · source default · version mapping unconfirmed
Meta · xhigh
| Compare | Input / outputUSD / 1M tokens | ||||||||
|---|---|---|---|---|---|---|---|---|---|
01 Claude Fable 5.1 ↗Anthropic · max | 79.6 | 56.3 | 94.42 sources | 83.4 | 1,501±6 | $200 | 1.00M | $10 / $50Direct API | |
02 Claude Opus 5.5 ↗Anthropic · max | 79.4 | 60.7 | — | 83.2 | — | $80 | 1.00M | $4 / $20Direct API | |
03 Muse Spark 1.3 ↗Meta · xhigh | 79.3 | 70.5 | — | 81.6 | — | $21 | 1.05M | $1.25 / $4.25OpenRouter catalog | |
04 Claude Fable 5 ↗Anthropic · source ID max; display xhighSource labels disagree | 79.0 | 55.8 | — | 83.0 | — | $200 | 1.00M | $10 / $50Direct API | |
05 DeepSeek V4.1 Flash ↗DeepSeek · maxOpen weights | 78.9 | 84.2 | — | 81.1 | — | $1.94 | 1.05M | $0.0158 / $0.891OpenRouter catalog | |
06 GPT-6 Astra ↗OpenAI · max | 78.6 | 55.6 | 44.42 sources | 82.2 | 1,475±7 | $200 | 1.05M | $10 / $50Direct API | |
07 Muse Wonder ↗Meta · max · version 1.0 | 78.3 | 60.0 | — | 81.2 | — | $80 | 1.05M | $4 / $20OpenRouter catalog | |
08 GPT-6.1 Sol ↗OpenAI · max | 77.7 | 64.5 | 55.62 sources | 81.6 | 1,484±9 | $40 | 1.05M | $2 / $10Direct API | |
09 Kimi K3 ↗Moonshot AI · source defaultOpen weights | 77.4 | 65.6 | — | 79.2 | — | $35 | 1.05M | $0.5 / $15OpenRouter catalog | |
10 GPT-5.6 Sol ↗OpenAI · source ID max; display xhighSource labels disagree | 77.1 | 59.1 | — | 81.1 | — | $80 | 1.05M | $4 / $20Direct API | |
11 Qwen3.8 Max ↗Alibaba · source default · version mapping unconfirmed | 77.0 | — | — | 78.5 | — | — | Not verified | Not verified | |
12 Muse Spark 1.2 ↗Meta · xhigh | 76.3 | 68.4 | 55.62 sources | 78.0 | 1,492±8 | $21 | 1.05M | $1.25 / $4.25OpenRouter catalog | |
13 Qwen3.8 Flash Next ↗Alibaba · source defaultOpen weights | 76.2 | — | — | 76.2 | — | — | Not verified | Not verified | |
14 Gemini 3.7 Flash ↗Google · high | 76.1 | 71.3 | 50.02 sources | 78.8 | 1,486±5 · preliminary | $15 | 1.05M | $0.75 / $3.75OpenRouter catalog | |
15 GPT-5.5 ↗OpenAI · xhigh | 75.8 | 54.6 | — | 80.2 | — | $110 | 1.05M | $5 / $30Direct API | |
16 Claude Opus 5 ↗Anthropic · max | 75.7 | 56.1 | 61.12 sources | 80.1 | 1,489±5 | $100 | 1.00M | $5 / $25Direct API | |
17 Grok 4.6 ↗xAI · source default | 75.3 | 65.6 | — | 78.0 | — | $32 | 500K | $2 / $6OpenRouter catalog | |
18 DeepSeek V4 Flash Vision Exp ↗DeepSeek · experimental · source defaultOpen weightsPreview | 75.1 | 79.5 | — | 76.8 | — | $3.45 | 1.05M | $0.2156 / $0.6468OpenRouter catalog | |
19 GPT-6 Sol ↗OpenAI · max | 74.7 | 62.4 | — | 79.2 | — | $40 | 1.05M | $2 / $10Direct API | |
20 GPT-5.4 ↗OpenAI · xhigh | 74.4 | 58.8 | — | 78.0 | — | $55 | 1.05M | $2.5 / $15Direct API | |
21 GPT-5.6 Terra ↗OpenAI · source ID max; display xhighSource labels disagree | 74.1 | 60.4 | — | 77.9 | — | $44 | 1.05M | $2 / $12Direct API | |
22 Muse Spark 1.1 ↗Meta · source ID xhigh; display highSource labels disagree | 74.0 | 66.8 | — | 75.3 | — | $21 | 1.05M | $1.25 / $4.25OpenRouter catalog | |
23 GLM-5.3 ↗Z.ai · source defaultOpen weights | 73.7 | 73.8 | — | 76.1 | — | $10.6 | 1.05M | $0.1 / $4.8OpenRouter catalog | |
24 Grok 4.7 ↗xAI · xhigh | 73.7 | 64.5 | — | 77.4 | — | $32 | 500K | $2 / $6OpenRouter catalog | |
25 Qwen3.8 27B ↗Alibaba · source defaultOpen weights | 73.5 | 76.3 | — | 75.3 | — | $6.8 | 1.00M | $0.24 / $2.2OpenRouter catalog | |
26 Gemini 3.8 Flash ↗Google · high | 73.3 | 69.4 | 55.62 sources | 75.8 | 1,497±5 · preliminary | $15 | 1.05M | $0.75 / $3.75OpenRouter catalog | |
27 Claude Sonnet 5 ↗Anthropic · xhigh | 73.3 | 61.4 | — | 76.0 | — | $40 | 1.00M | $2 / $10Direct API | |
28 DeepSeek V4 Pro 0813 ↗DeepSeek · source defaultOpen weights | 73.3 | 74.0 | — | 77.4 | — | $10.56 | 1.05M | $0.66 / $1.98OpenRouter catalog | |
29 Gemini 3.1 Pro Preview ↗Google · highPreview | 73.2 | 59.8 | — | 77.0 | — | $44 | 1.05M | $2 / $12OpenRouter catalog | |
30 Grok 4.5 ↗xAI · source default | 73.1 | 64.1 | — | 75.8 | — | $32 | 500K | $2 / $6OpenRouter catalog | |
31 Claude Opus 4.8 ↗Anthropic · max | 73.0 | 54.2 | — | 76.2 | — | $100 | 1.00M | $5 / $25Direct API | |
32 Claude Opus 4.7 ↗Anthropic · xhigh | 72.9 | 54.1 | — | 76.5 | — | $100 | 1.00M | $5 / $25Direct API | |
33 Mistral Large 4 ↗Mistral · highPreview | 72.0 | 71.6 | — | 71.8 | — | $10.98 | 1.05M | $0.68 / $2.09OpenRouter catalog | |
34 DeepSeek V4 Flash 0731 ↗DeepSeek · source defaultOpen weights | 71.1 | 79.3 | — | 74.2 | — | $1.44 | 1.05M | $0.016 / $0.64OpenRouter catalog | |
35 Claude Sonnet 5.5 ↗Anthropic · max | 71.0 | 59.8 | — | 75.7 | — | $40 | 1.00M | $2 / $10Direct API | |
36 Gemini 3.5 Flash ↗Google · high | 70.8 | 61.5 | 16.72 sources | 74.6 | 1,478±4 | $33 | 1.05M | $1.5 / $9OpenRouter catalog | |
37 Claude Opus 4.6 ↗Anthropic · high · adaptive thinking | 70.5 | 52.5 | 55.62 sources | 74.5 | 1,504±3 | $100 | 1.00M | $5 / $25Direct API | |
38 Qwen3.7 Max ↗Alibaba · source default | 70.4 | 63.3 | — | 73.1 | — | $23.6 | 1.00M | $1.48 / $4.43OpenRouter catalog | |
39 GPT-5.6 Luna ↗OpenAI · source ID max; display xhighSource labels disagree | 70.4 | 75.7 | — | 73.6 | — | $4.4 | 1.05M | $0.2 / $1.2Direct API | |
40 Gemini 3.6 Flash ↗Google · high | 70.3 | 67.3 | 11.12 sources | 73.6 | 1,482±4 | $15 | 1.05M | $0.75 / $3.75OpenRouter catalog | |
41 GPT-5.2 Codex ↗OpenAI · source default | 69.9 | 56.4 | — | 74.0 | — | $45.5 | 400K | $1.75 / $14OpenRouter catalog | |
42 GPT-5.2 ↗OpenAI · high · 2025-12-11 | 69.8 | 56.4 | — | 74.6 | — | $45.5 | 400K | $1.75 / $14OpenRouter catalog | |
43 Claude Sonnet 4.6 ↗Anthropic · medium · adaptive thinking | 69.4 | 54.8 | — | 73.0 | — | $60 | 1.00M | $3 / $15Direct API | |
44 Inkling ↗Thinking Machines · xhighOpen weights | 68.9 | 64.3 | — | 71.9 | — | $18.1 | 524K | $1 / $4.05OpenRouter catalog | |
45 GLM-5.2 ↗Z.ai · source defaultOpen weights | 68.5 | 67.1 | — | 73.2 | — | $14.4 | 1.05M | $0.04 / $7OpenRouter catalog | |
46 GPT-5.4 nano ↗OpenAI · xhigh | 67.7 | 73.3 | — | 69.6 | — | $4.5 | 400K | $0.2 / $1.25Direct API | |
47 GPT-6 Luna ↗OpenAI · max | 67.7 | 75.5 | — | 72.0 | — | $2 | 1.05M | $0.1 / $0.5Direct API | |
48 GLM-5.3 Flash ↗Z.ai · source defaultOpen weights | 67.2 | 74.5 | — | 71.6 | — | $2.5 | 1.05M | $0.15 / $0.5OpenRouter catalog | |
49 DeepSeek V4 Pro ↗DeepSeek · source defaultOpen weights | 67.1 | 67.1 | — | 71.6 | — | $13.37 | 1.05M | $0.9553 / $1.91OpenRouter catalog | |
50 Kimi K2.6 ↗Moonshot AI · thinkingOpen weights | 66.9 | 67.5 | — | 70.5 | — | $13.32 | 262K | $0.65 / $3.41OpenRouter catalog | |
51 Claude Opus 4.5 ↗Anthropic · high · 64K thinking | 66.7 | 49.8 | — | 72.6 | — | $100 | 200K | $5 / $25Direct API | |
52 Claude Haiku 5.5 ↗Anthropic · max | 66.0 | 74.4 | — | 67.9 | — | $2 | 1.00M | $0.1 / $0.5Direct API | |
53 Kimi K2.7 Code ↗Moonshot AI · source defaultOpen weights | 64.8 | 65.0 | — | 68.4 | — | $13.41 | 262K | $0.6712 / $3.35OpenRouter catalog | |
54 Qwen3.6 Plus ↗Alibaba · source default | 63.9 | 69.0 | — | 68.9 | — | $7.15 | 1.00M | $0.325 / $1.95OpenRouter catalog | |
55 Nemotron 3 Ultra ↗NVIDIA · 550B A55B · source defaultOpen weights | 63.8 | 66.4 | — | 67.4 | — | $10.8 | 262K | $0.6 / $2.4OpenRouter catalog | |
56 MiniMax M3 ↗MiniMax · source default | 63.1 | 69.5 | — | 67.3 | — | $5.4 | 1.05M | $0.3 / $1.2OpenRouter catalog | |
57 GPT-5.4 mini ↗OpenAI · xhigh | 62.5 | 60.8 | — | 66.4 | — | $16.5 | 400K | $0.75 / $4.5Direct API | |
58 DeepSeek V4 Flash ↗DeepSeek · source defaultOpen weights | 61.6 | 73.1 | — | 65.5 | — | $1.27 | 1.05M | $0.091 / $0.182OpenRouter catalog | |
59 Qwen3.6 27B ↗Alibaba · source defaultOpen weights | 60.0 | 65.3 | — | 64.0 | — | $9.6 | 262K | $0.32 / $3.2OpenRouter catalog | |
60 Gemini 3.5 Flash-Lite ↗Google · high | 59.5 | 65.5 | — | 63.9 | — | $8 | 1.05M | $0.3 / $2.5OpenRouter catalog | |
61 Grok 4.3 ↗xAI · source default | 56.0 | 55.8 | — | 62.2 | — | $17.5 | 1.00M | $1.25 / $2.5OpenRouter catalog |
— means a comparable result is unavailable. Consensus combines results for only 10 matched configurations. Model IDs, reasoning settings, source links and pricing differences are available on every model page. Downloadable model files can still have restrictions on commercial use.
LOOK BEYOND ONE NUMBER
| Measure | Claude Opus 5.5max | GPT-6 Solmax | DeepSeek V4.1 Flashmax |
|---|---|---|---|
| Agent Fit | 79.4 | 74.7 | 78.9 |
| Practical Value | 60.7 | 62.4 | 84.2 |
| Evidence Consensus | Insufficient evidence | Insufficient evidence | Insufficient evidence |
| Workload cost | $80 | $40 | $1.94 |
| Context window | 1.00M | 1.05M | 1.05M |
| Tool calling | Supported | Supported | Supported |
| Reasoning | 92.2 | 88.7 | 86.7 |
| Coding | 89.3 | 81.8 | 80.0 |
| Agentic coding | 71.7 | 52.9 | 77.3 |
| Mathematics | 97.1 | 96.4 | 93.3 |
| Data analysis | 80.3 | 81.2 | 79.3 |
| Language | 86.3 | 85.3 | 81.2 |
| Instruction following | 65.7 | 68.6 | 70.0 |
| Input / output per 1M | $4 / $20 | $2 / $10 | $0.0158 / $0.891 |