LLM Price vs Performance Chart
As of November 18, 2026, Gemini 4 Carbon (Google) is the cheapest model within 90% of the overall leader's score: 85 points at $10.00 per million output tokens, 75% less than Claude Fable 5.2 (92.8 points, $40.00).
Compare cost vs performance for ranked LLMs: each model’s active public score against its current API price. The shortlist favors the cheapest model that keeps at least 90% of the leader’s score; the efficiency frontier shows every undominated tradeoff.
- 90% pick, overallGemini 4 CarbonGoogle85$10.00 per 1M output tokens75% below the leader’s price
- Highest score, overallClaude Fable 5.2Anthropic92.8$40.00 per 1M output tokensAvailability restrictions apply
- Price floor, overallQwen3.7 FlashAlibaba46.6$0.13 per 1M output tokensLowest price in this view
Best model for every budget
The highest overall score in each price band, with the next model in the band. Bands apply to the output price per million tokens.
The share compares each band’s top score with the highest score at any price.
| Budget per 1M output tokens | Top scorer |
|---|---|
| Under $1 | Claude Haiku 5.562.9 points · $0.50 · 68% of the leaderNext: MiniMax M3.1 Flash (58.2) |
| $1 to $5 | Gemini 3.8 Flash64.4 points · $3.75 · 69% of the leaderNext: Gemini 3.7 Flash (64.4) |
| $5 to $15 | Gemini 4 Carbon85.0 points · $10.00 · 92% of the leaderNext: Claude Sonnet 5.5 (80.6) |
| $15 and over | Claude Fable 5.292.8 points · $40.00 · 100% of the leaderNext: Claude Opus 5.5 (85.8) |
Score against price
Each point is a model, with price on the horizontal axis and score on the vertical, so the chart shows cost vs performance for every priced model in the view. The dark line is the efficiency frontier; for anything in the gray area under it, another model costs no more and scores at least as high. Prices use a log scale. Open a point for its details.
96 priced models match these filters. The blended view assumes three input tokens for every output token.
Select a point, or tab to it and press Enter, to see its price, its status and its nearest better value, the closest in price of the models that cost no more and score at least as high.
Best value by category
For each category with enough priced models, the leader, the cheapest model within 90% of its score, and what the pick saves. Output prices, one row per model family.
Select a category to show it on the chart.
| Category | 90% pick |
|---|---|
| OverallLeader: Claude Fable 5.2 (92.8, $40.00) | Gemini 4 Carbon85.0 points · $10.00 · saves 75% |
| AgenticLeader: Claude Fable 5.2 (90.8, $40.00) | Gemini 4 Carbon85.1 points · $10.00 · saves 75% |
| CodingLeader: Claude Fable 5.2 (89.7, $40.00) | Gemini 4 Carbon83.4 points · $10.00 · saves 75% |
| ReasoningLeader: Claude Fable 5.2 (91.0, $40.00) | Gemini 4 Carbon84.0 points · $10.00 · saves 75% |
| KnowledgeLeader: Claude Fable 5.2 (98.5, $40.00) | Claude Opus 5.589.5 points · $20.00 · saves 50% |
| MultimodalLeader: Claude Fable 5.2 (89.9, $40.00) | Claude Opus 5.584.2 points · $20.00 · saves 50% |
| Instruction FollowingLeader: Grok 4.3 (97.1, $2.50) | MiniMax M393.8 points · $1.20 · saves 52% |
All priced models
Every model on the chart, under the same filters, ranked by overall score. The status says whether another model costs no more and scores at least as high; the last column, nearest better value, names the closest such model in price.
96 models · 10 on the frontier
| # | Model | Nearest better value | ||||
|---|---|---|---|---|---|---|
| 1 | Claude Fable 5.2Anthropic · Estimated | 92.8 | $40.00 | $16.00 | On the frontier | None |
| 2 | Claude Opus 5.5Anthropic · Supported | 85.8 | $20.00 | $8.00 | On the frontier | None |
| 3 | Gemini 4 CarbonGoogle · Estimated | 85.0 | $10.00 | $4.00 | On the frontier | None |
| 4 | GPT-6 AstraOpenAI · Supported | 83.0 | $50.00 | $20.00 | 3 cheaper models score at least as high | Claude Fable 5.2 $40.00 |
| 5 | Claude Sonnet 5.5Anthropic · Supported | 80.6 | $10.00 | $4.00 | 1 model at the same price scores higher | Gemini 4 Carbon $10.00 |
| 6 | Claude Opus 5Anthropic · Supported | 76.0 | $25.00 | $10.00 | 3 cheaper models score at least as high | Claude Opus 5.5 $20.00 |
| 7 | Kimi K3.1 ProMoonshot AI · Estimated | 71.3 | $10.00 | $4.00 | 2 models at the same price score higher | Gemini 4 Carbon $10.00 |
| 8 | GPT-5.6 SolOpenAI · Supported | 70.9 | $20.00 | $8.00 | 4 models score at least as high for the same price or less | Claude Opus 5.5 $20.00 |
| 9 | Kimi K3Moonshot AI · Supported | 66.7 | $15.00 | $6.00 | 3 cheaper models score at least as high | Gemini 4 Carbon $10.00 |
| 10 | Muse WonderMeta · Estimated | 66.1 | $20.00 | $8.00 | 6 models score at least as high for the same price or less | Claude Opus 5.5 $20.00 |
| 11 | Claude Opus 4.8Anthropic · Supported | 65.9 | $25.00 | $10.00 | 8 models score at least as high for the same price or less | Claude Opus 5 $25.00 |
| 12 | GPT-5.5OpenAI · Supported | 65.9 | $30.00 | $11.25 | 9 cheaper models score at least as high | Claude Opus 5 $25.00 |
| 13 | Claude Opus 4.7Anthropic · Supported | 65.6 | $25.00 | $10.00 | 9 models score at least as high for the same price or less | Claude Opus 5 $25.00 |
| 14 | Grok 4.6xAI · Supported | 64.7 | $6.00 | $3.00 | On the frontier | None |
| 15 | Gemini 3.8 FlashGoogle · Supported | 64.4 | $3.75 | $1.50 | On the frontier | None |
| 16 | Gemini 3.7 FlashGoogle · Supported | 64.4 | $3.75 | $1.50 | On the frontier | None |
| 17 | Qwen4 MaxAlibaba · Estimated | 64.4 | $4.50 | $2.25 | 2 cheaper models score at least as high | Gemini 3.8 Flash $3.75 |
| 18 | GPT-5.4OpenAI · Supported | 64.1 | $15.00 | $5.63 | 8 models score at least as high for the same price or less | Kimi K3 $15.00 |
| 19 | DeepSeek V4.1 Flash 1017DeepSeek · Supported | 63.9 | $1.20 | $0.52 | On the frontier | None |
| 20 | Muse Spark 1.2Meta · Supported | 63.7 | $4.25 | $2.00 | 3 cheaper models score at least as high | Gemini 3.8 Flash $3.75 |
| 21 | Claude Haiku 5.5Anthropic · Estimated | 62.9 | $0.50 | $0.20 | On the frontier | None |
| 22 | Gemini 3.1 ProGoogle · Estimated | 62.9 | $12.00 | $4.50 | 10 cheaper models score at least as high | Gemini 4 Carbon $10.00 |
| 23 | DeepSeek V4 ProDeepSeek · Estimated | 62.7 | $3.96 | $1.98 | 4 cheaper models score at least as high | Gemini 3.8 Flash $3.75 |
| 24 | Claude Sonnet 5Anthropic · Supported | 62.3 | $10.00 | $4.00 | 11 models score at least as high for the same price or less | Gemini 4 Carbon $10.00 |
| 25 | GPT-5.3 CodexOpenAI · Estimated | 62.2 | $14.00 | $4.81 | 13 cheaper models score at least as high | Gemini 3.1 Pro $12.00 |
| 26 | Gemini 3.5 FlashGoogle · Supported | 61.8 | $9.00 | $3.38 | 8 cheaper models score at least as high | Grok 4.6 $6.00 |
| 27 | Grok 4.5xAI · Supported | 61.5 | $6.00 | $3.00 | 8 models score at least as high for the same price or less | Grok 4.6 $6.00 |
| 28 | Claude Opus 4.6Anthropic · Supported | 61.3 | $25.00 | $10.00 | 24 models score at least as high for the same price or less | Claude Opus 5 $25.00 |
| 29 | Gemini 3.6 FlashGoogle · Supported | 61.2 | $3.75 | $1.50 | 4 models score at least as high for the same price or less | Gemini 3.8 Flash $3.75 |
| 30 | GLM-5.2Z.AI · Supported | 58.5 | $4.40 | $2.15 | 7 cheaper models score at least as high | Muse Spark 1.2 $4.25 |
| 31 | MiniMax M3.1 FlashMiniMax · Estimated | 58.2 | $0.60 | $0.26 | 1 cheaper model scores at least as high | Claude Haiku 5.5 $0.50 |
| 32 | GPT-5.2OpenAI · Supported | 57.8 | $14.00 | $4.81 | 19 models score at least as high for the same price or less | GPT-5.3 Codex $14.00 |
| 33 | Gemini 3 ProGoogle · Supported | 57.3 | $12.00 | $4.50 | 18 models score at least as high for the same price or less | Gemini 3.1 Pro $12.00 |
| 34 | MiMo-V2.6-ProXiaomi · Estimated | 56.7 | $0.87 | $0.54 | 2 cheaper models score at least as high | MiniMax M3.1 Flash $0.60 |
| 35 | GPT-5.1OpenAI · Supported | 56.1 | $10.00 | $3.44 | 18 models score at least as high for the same price or less | Gemini 4 Carbon $10.00 |
| 36 | Kimi K2.6Moonshot AI · Estimated | 55.8 | $4.00 | $1.71 | 8 cheaper models score at least as high | DeepSeek V4 Pro $3.96 |
| 37 | Claude Sonnet 4.6Anthropic · Supported | 55.8 | $15.00 | $6.00 | 26 models score at least as high for the same price or less | Kimi K3 $15.00 |
| 38 | Step 5 PreviewStepFun · Estimated | 54.8 | $2.70 | $1.43 | 4 cheaper models score at least as high | DeepSeek V4.1 Flash 1017 $1.20 |
| 39 | Grok 4.20xAI · Estimated | 54.5 | $2.50 | $1.56 | 4 cheaper models score at least as high | DeepSeek V4.1 Flash 1017 $1.20 |
| 40 | Qwen3.5 397BAlibaba · Estimated | 54.3 | $3.60 | $1.35 | 6 cheaper models score at least as high | Step 5 Preview $2.70 |
| 41 | GPT-5.2-CodexOpenAI · Estimated | 54.2 | $14.00 | $4.81 | 27 models score at least as high for the same price or less | GPT-5.3 Codex $14.00 |
| 42 | Gemini 3 FlashGoogle · Supported | 54.0 | $3.00 | $1.13 | 6 cheaper models score at least as high | Step 5 Preview $2.70 |
| 43 | Muse EmberMeta · Estimated | 53.8 | $0.50 | $0.20 | 1 model at the same price scores higher | Claude Haiku 5.5 $0.50 |
| 44 | Kimi K2.7 CodeMoonshot AI · Estimated | 53.4 | $4.00 | $1.71 | 14 models score at least as high for the same price or less | Kimi K2.6 $4.00 |
| 45 | Claude Opus 4.5Anthropic · Supported | 52.0 | $25.00 | $10.00 | 41 models score at least as high for the same price or less | Claude Opus 5 $25.00 |
| 46 | Grok 4.3xAI · Supported | 51.7 | $2.50 | $1.56 | 6 models score at least as high for the same price or less | Grok 4.20 $2.50 |
| 47 | Kimi K2.5Moonshot AI · Supported | 51.7 | $3.00 | $1.20 | 9 models score at least as high for the same price or less | Gemini 3 Flash $3.00 |
| 48 | Mistral Large 4Mistral · Supported | 51.4 | $2.10 | $1.05 | 5 cheaper models score at least as high | DeepSeek V4.1 Flash 1017 $1.20 |
| 49 | DeepSeek V3.2DeepSeek · Supported | 49.1 | $0.42 | $0.32 | On the frontier | None |
| 50 | Quasar 438BMultiverse Computing · Estimated | 49.1 | $1.80 | $0.90 | 6 cheaper models score at least as high | DeepSeek V4.1 Flash 1017 $1.20 |
| 51 | GPT-5OpenAI · Estimated | 49.0 | $10.00 | $3.44 | 31 models score at least as high for the same price or less | Gemini 4 Carbon $10.00 |
| 52 | GPT-5.1-CodexOpenAI · Estimated | 48.9 | $10.00 | $3.44 | 32 models score at least as high for the same price or less | Gemini 4 Carbon $10.00 |
| 53 | MiniMax M3MiniMax · Supported | 48.4 | $1.20 | $0.52 | 6 models score at least as high for the same price or less | DeepSeek V4.1 Flash 1017 $1.20 |
| 54 | Qwen3.5 PlusAlibaba · Estimated | 48.4 | $2.40 | $0.90 | 9 cheaper models score at least as high | Mistral Large 4 $2.10 |
| 55 | MiniMax M2.5MiniMax · Supported | 48.3 | $1.20 | $0.52 | 7 models score at least as high for the same price or less | DeepSeek V4.1 Flash 1017 $1.20 |
| 56 | Gemini 2.5 ProGoogle · Supported | 48.1 | $10.00 | $3.44 | 36 models score at least as high for the same price or less | Gemini 4 Carbon $10.00 |
| 57 | InklingThinking Machines Lab · Supported | 47.7 | $4.68 | $2.57 | 26 cheaper models score at least as high | Qwen4 Max $4.50 |
| 58 | Gemini 3.5 Flash-LiteGoogle · Supported | 47.1 | $2.50 | $0.85 | 13 models score at least as high for the same price or less | Grok 4.20 $2.50 |
| 59 | Qwen3.7 FlashAlibaba · Estimated | 46.6 | $0.13 | $0.06 | On the frontier | None |
| 60 | o3OpenAI · Supported | 46.4 | $8.00 | $3.50 | 31 cheaper models score at least as high | Grok 4.6 $6.00 |
| 61 | MiniMax M2.7MiniMax · Supported | 46.1 | $1.20 | $0.52 | 9 models score at least as high for the same price or less | DeepSeek V4.1 Flash 1017 $1.20 |
| 62 | Claude Sonnet 4.5Anthropic · Estimated | 46.0 | $15.00 | $6.00 | 50 models score at least as high for the same price or less | Kimi K3 $15.00 |
| 63 | Gemini 3.1 Flash-LiteGoogle · Estimated | 45.8 | $1.50 | $0.56 | 10 cheaper models score at least as high | DeepSeek V4.1 Flash 1017 $1.20 |
| 64 | Step 3.7 FlashStepFun · Estimated | 44.4 | $1.15 | $0.44 | 6 cheaper models score at least as high | MiMo-V2.6-Pro $0.87 |
| 65 | Qwen3.5 FlashAlibaba · Estimated | 44.3 | $0.40 | $0.18 | 1 cheaper model scores at least as high | Qwen3.7 Flash $0.13 |
| 66 | Grok 3 MinixAI · Estimated | 43.4 | $0.50 | $0.35 | 5 models score at least as high for the same price or less | Claude Haiku 5.5 $0.50 |
| 67 | Trinity-Large-ThinkingArcee AI · Estimated | 42.8 | $0.90 | $0.41 | 8 cheaper models score at least as high | MiMo-V2.6-Pro $0.87 |
| 68 | Step 3.5 FlashStepFun · Estimated | 41.3 | $0.30 | $0.15 | 1 cheaper model scores at least as high | Qwen3.7 Flash $0.13 |
| 69 | Gemini 2.5 FlashGoogle · Supported | 41.3 | $2.50 | $0.85 | 22 models score at least as high for the same price or less | Grok 4.20 $2.50 |
| 70 | DeepSeek-R1DeepSeek · Supported | 40.1 | $2.19 | $0.96 | 18 cheaper models score at least as high | Mistral Large 4 $2.10 |
| 71 | Claude Haiku 4.5Anthropic · Estimated | 39.8 | $5.00 | $2.00 | 38 cheaper models score at least as high | Inkling $4.68 |
| 72 | Claude 4.1 OpusAnthropic · Estimated | 39.1 | $75.00 | $30.00 | 71 cheaper models score at least as high | GPT-6 Astra $50.00 |
| 73 | Claude 4 SonnetAnthropic · Estimated | 37.3 | $15.00 | $6.00 | 60 models score at least as high for the same price or less | Kimi K3 $15.00 |
| 74 | GPT-4.1OpenAI · Supported | 35.6 | $8.00 | $3.50 | 42 models score at least as high for the same price or less | o3 $8.00 |
| 75 | Grok 4.1 FastxAI · Estimated | 35.5 | $0.50 | $0.28 | 7 models score at least as high for the same price or less | Claude Haiku 5.5 $0.50 |
| 76 | o1OpenAI · Estimated | 35.2 | $60.00 | $26.25 | 74 cheaper models score at least as high | GPT-6 Astra $50.00 |
| 77 | GLM-4.5-AirZ.AI · Estimated | 34.2 | $1.10 | $0.43 | 11 cheaper models score at least as high | Trinity-Large-Thinking $0.90 |
| 78 | LFM2.5-24B-A2BLiquidAI · Estimated | 33.9 | $1.00 | $0.33 | 11 cheaper models score at least as high | Trinity-Large-Thinking $0.90 |
| 79 | Mistral Large 3Mistral · Estimated | 33.2 | $1.50 | $0.75 | 19 models score at least as high for the same price or less | Gemini 3.1 Flash-Lite $1.50 |
| 80 | Grok Code Fast 1xAI · Estimated | 33.1 | $1.50 | $0.53 | 20 models score at least as high for the same price or less | Gemini 3.1 Flash-Lite $1.50 |
| 81 | Kimi K2Moonshot AI · Estimated | 32.7 | $2.50 | $1.07 | 29 models score at least as high for the same price or less | Grok 4.20 $2.50 |
| 82 | Mercury 2.5Inception · Estimated | 32.6 | $0.15 | $0.07 | 1 cheaper model scores at least as high | Qwen3.7 Flash $0.13 |
| 83 | Mercury 2Inception · Estimated | 32.4 | $0.75 | $0.38 | 10 cheaper models score at least as high | MiniMax M3.1 Flash $0.60 |
| 84 | DeepSeek V3DeepSeek · Supported | 31.1 | $1.10 | $0.48 | 15 models score at least as high for the same price or less | GLM-4.5-Air $1.10 |
| 85 | Solar Pro 3Upstage · Estimated | 29.5 | $0.60 | $0.26 | 10 models score at least as high for the same price or less | MiniMax M3.1 Flash $0.60 |
| 86 | Mistral Medium 3.5 128BMistral · Estimated | 28.2 | $7.50 | $3.00 | 51 cheaper models score at least as high | Grok 4.6 $6.00 |
| 87 | Command A+Cohere · Estimated | 28.2 | $10.00 | $4.38 | 63 models score at least as high for the same price or less | Gemini 4 Carbon $10.00 |
| 88 | Mistral Medium 3Mistral · Estimated | 28.1 | $2.00 | $0.80 | 26 cheaper models score at least as high | Quasar 438B $1.80 |
| 89 | Gemini 1.5 ProGoogle · Estimated | 27.8 | $5.00 | $2.19 | 50 models score at least as high for the same price or less | Claude Haiku 4.5 $5.00 |
| 90 | Claude 3 OpusAnthropic · Supported | 27.6 | $75.00 | $30.00 | 89 models score at least as high for the same price or less | Claude 4.1 Opus $75.00 |
| 91 | Celeris-1Celeris · Estimated | 27.0 | $0.70 | $0.33 | 11 cheaper models score at least as high | MiniMax M3.1 Flash $0.60 |
| 92 | GPT-4oOpenAI · Estimated | 26.0 | $10.00 | $4.38 | 67 models score at least as high for the same price or less | Gemini 4 Carbon $10.00 |
| 93 | Mistral Small 4Mistral · Estimated | 25.7 | $0.60 | $0.26 | 11 models score at least as high for the same price or less | MiniMax M3.1 Flash $0.60 |
| 94 | Claude 3.5 SonnetAnthropic · Estimated | 24.8 | $15.00 | $6.00 | 79 models score at least as high for the same price or less | Kimi K3 $15.00 |
| 95 | GPT-4 TurboOpenAI · Supported | 20.8 | $30.00 | $15.00 | 89 models score at least as high for the same price or less | GPT-5.5 $30.00 |
| 96 | Claude 3 HaikuAnthropic · Supported | 13.3 | $1.25 | $0.50 | 24 cheaper models score at least as high | DeepSeek V4.1 Flash 1017 $1.20 |
Prices per million tokens from the pricing catalog, verified November 18, 2026. Blended assumes three input tokens for every output token.
How to read this page
Two rules decide what the page calls good price to performance. Neither turns benchmark points into dollars.
- Efficiency frontier
- The models no other model beats on both price and score. For every model off the frontier, at least one other model costs no more and scores at least as high.
- The 90% rule
- The pick is the cheapest model that keeps at least 90% of the leader’s score. Score per dollar alone favors the cheapest weak models, because benchmark points are an index, not units of finished work.
Questions
Which LLM API is cheapest?
Among ranked models with a published API price, Qwen3.7 Flash (Alibaba) has the lowest output price: $0.13 per million output tokens, with an overall score of 46.6. On the 3:1 blended basis the cheapest is Qwen3.7 Flash at $0.06 per million tokens. The cheapest model is rarely the one to pick, because its score is often far below the leader's.
What is the cheapest LLM that is still good?
That depends on the score you need, so set a floor and read the budget table. Under $1 per million output tokens, the top overall score is Claude Haiku 5.5 at 62.9, 68% of the leader's. The 90% pick in the hero is one floor; the frontier rows in the full table show the others.
Which LLM has the best price to performance?
The 90% pick at the top of this page, if you want at least 90% of the overall leader’s score at the lowest price. For a different score floor, use the efficiency frontier: its 10 models in the overall view are the cost-efficient choices, since no other model beats any of them on both price and score.
What is the efficiency frontier?
The efficiency frontier is the set of models that no other model beats on both price and score. For a model below and to the right of it, another model costs no more and scores at least as high. In the overall view, 10 models are on it, from Qwen3.7 Flash at $0.13 to Claude Fable 5.2 at $40.00 per million output tokens.
Should I compare output price or blended price?
Output price is the default because output tokens cost more and dominate the bill for long answers and agent loops. The 3:1 blended price weighs three input tokens for every output token, which fits chat and retrieval work with long prompts. On blended price the overall 90% pick stays the same.
How are scores calculated?
The chart reads the active public ranking lane for each category and pairs that score with current catalog pricing. Overall, coding, and agentic views use the current BenchAlign methodology when enabled; other categories use the provisional composite. The 90% shortlist is a decision aid, not a claim that benchmark points convert directly into dollars.
Inside one provider the tradeoff is easiest to see: Gemini 4 Carbon or Gemini 4 Hydrogen is a price-for-capability decision within Google's own line.
Keep the value shortlist current
One weekly email when price changes or new benchmark evidence alter the models worth shortlisting.
Read a sample issueJoin 7,000+ subscribers.
One email each week. Unsubscribe anytime.