LLM Price vs. Performance Benchmark
Compare real-time LLM API pricing against verified benchmark scores. Track Pareto frontier models to identify the optimal balance of intelligence and cost for your workload.
Filter Models & Data Parameters
Customize the score lane, pricing range, vendor filters to update the scatter plot and data tables below.
Price–Performance Pareto Frontier
Models on the dotted line represent the best performance available at their given price point (Pareto frontier). Click any point to inspect exact scores and token costs.
Scatter plot with 54 models. 8 Pareto frontier points. Score values range from 40.7 to 83.4. Selected cost values range from 0.3 to 168.0.
Model Performance & Value Leaderboard
Compare efficiency metrics, including score-per-dollar values, across all current models.
| Model | Score | Selected cost | Score / dollar | Evidence | Frontier |
|---|---|---|---|---|---|
| MiniMax M3 | 68.7 | $1.2 / 1M output price | 57.28 | Supported | Pareto frontier |
| GLM-5.1 | 67.0 | $4.4 / 1M output price | 15.24 | Supported | Not on Pareto frontier |
| GPT-5.4 nano | 66.8 | $1.25 / 1M output price | 53.42 | Supported | Not on Pareto frontier |
| GLM-5-Turbo | 66.3 | $4 / 1M output price | 16.58 | Supported | Not on Pareto frontier |
| GLM-5 | 65.9 | $3.2 / 1M output price | 20.59 | Supported | Not on Pareto frontier |
| Gemini 3.5 Flash-Lite | 65.5 | $2.5 / 1M output price | 26.2 | Supported | Not on Pareto frontier |
| Grok 4.3 | 64.3 | $2.5 / 1M output price | 25.7 | Supported | Not on Pareto frontier |
| Inkling-Small | 64.0 | $1.44 / 1M output price | 44.46 | Supported | Not on Pareto frontier |
| MiniMax M2.7 | 63.3 | $1.2 / 1M output price | 52.74 | Supported | Not on Pareto frontier |
| GLM-5V-Turbo | 62.5 | $4 / 1M output price | 15.63 | Supported | Not on Pareto frontier |
MiniMax M3
- Score
- 68.7
- Selected cost
- $1.2 / 1M output price
- Score per dollar
- 57.28
- Evidence
- Supported
- Frontier
- Pareto frontier
GLM-5.1
- Score
- 67.0
- Selected cost
- $4.4 / 1M output price
- Score per dollar
- 15.24
- Evidence
- Supported
- Frontier
- Not on Pareto frontier
GPT-5.4 nano
- Score
- 66.8
- Selected cost
- $1.25 / 1M output price
- Score per dollar
- 53.42
- Evidence
- Supported
- Frontier
- Not on Pareto frontier
GLM-5-Turbo
- Score
- 66.3
- Selected cost
- $4 / 1M output price
- Score per dollar
- 16.58
- Evidence
- Supported
- Frontier
- Not on Pareto frontier
GLM-5
- Score
- 65.9
- Selected cost
- $3.2 / 1M output price
- Score per dollar
- 20.59
- Evidence
- Supported
- Frontier
- Not on Pareto frontier
Gemini 3.5 Flash-Lite
- Score
- 65.5
- Selected cost
- $2.5 / 1M output price
- Score per dollar
- 26.2
- Evidence
- Supported
- Frontier
- Not on Pareto frontier
Grok 4.3
- Score
- 64.3
- Selected cost
- $2.5 / 1M output price
- Score per dollar
- 25.7
- Evidence
- Supported
- Frontier
- Not on Pareto frontier
Inkling-Small
- Score
- 64.0
- Selected cost
- $1.44 / 1M output price
- Score per dollar
- 44.46
- Evidence
- Supported
- Frontier
- Not on Pareto frontier
MiniMax M2.7
- Score
- 63.3
- Selected cost
- $1.2 / 1M output price
- Score per dollar
- 52.74
- Evidence
- Supported
- Frontier
- Not on Pareto frontier
GLM-5V-Turbo
- Score
- 62.5
- Selected cost
- $4 / 1M output price
- Score per dollar
- 15.63
- Evidence
- Supported
- Frontier
- Not on Pareto frontier
View all 54 filtered models
| Model | Score | Selected cost | Score / dollar | Evidence | Frontier |
|---|---|---|---|---|---|
| Step 3.5 Flash | 54.6 | $0.3 / 1M output price | 181.83 | Supported | Pareto frontier |
| Qwen3.5 Flash | 47.7 | $0.4 / 1M output price | 119.12 | Supported | Not on Pareto frontier |
| DeepSeek V3.2 | 55.0 | $0.42 / 1M output price | 130.86 | Supported | Pareto frontier |
| Grok 4.1 Fast | 50.9 | $0.5 / 1M output price | 101.76 | Supported | Not on Pareto frontier |
| Mercury 2 | 48.4 | $0.75 / 1M output price | 64.55 | Supported | Not on Pareto frontier |
| Trinity-Large-Thinking | 48.2 | $0.9 / 1M output price | 53.52 | Supported | Not on Pareto frontier |
| GLM-4.5-Air | 47.3 | $1.1 / 1M output price | 43 | Supported | Not on Pareto frontier |
| DeepSeek V3 | 44.6 | $1.1 / 1M output price | 40.52 | Supported | Not on Pareto frontier |
| MiniMax M3 | 68.7 | $1.2 / 1M output price | 57.28 | Supported | Pareto frontier |
| MiniMax M2.7 | 63.3 | $1.2 / 1M output price | 52.74 | Supported | Not on Pareto frontier |
| MiniMax M2.5 | 58.9 | $1.2 / 1M output price | 49.1 | Supported | Not on Pareto frontier |
| GPT-5.4 nano | 66.8 | $1.25 / 1M output price | 53.42 | Supported | Not on Pareto frontier |
| Inkling-Small | 64.0 | $1.44 / 1M output price | 44.46 | Supported | Not on Pareto frontier |
| Gemini 3.1 Flash-Lite | 50.9 | $1.5 / 1M output price | 33.93 | Supported | Not on Pareto frontier |
| Mistral Large 3 | 49.8 | $1.5 / 1M output price | 33.21 | Supported | Not on Pareto frontier |
| GPT-5 mini | 43.3 | $2 / 1M output price | 21.65 | Supported | Not on Pareto frontier |
| DeepSeek-R1 | 51.2 | $2.19 / 1M output price | 23.39 | Supported | Not on Pareto frontier |
| Gemini 3.5 Flash-Lite | 65.5 | $2.5 / 1M output price | 26.2 | Supported | Not on Pareto frontier |
| Grok 4.3 | 64.3 | $2.5 / 1M output price | 25.7 | Supported | Not on Pareto frontier |
| Gemini 2.5 Flash | 47.8 | $2.5 / 1M output price | 19.1 | Supported | Not on Pareto frontier |
| Gemini 3 Flash | 60.1 | $3 / 1M output price | 20.03 | Supported | Not on Pareto frontier |
| Kimi K2.5 | 59.1 | $3 / 1M output price | 19.71 | Supported | Not on Pareto frontier |
| GLM-5 | 65.9 | $3.2 / 1M output price | 20.59 | Supported | Not on Pareto frontier |
| GLM-5-Turbo | 66.3 | $4 / 1M output price | 16.58 | Supported | Not on Pareto frontier |
| GLM-5V-Turbo | 62.5 | $4 / 1M output price | 15.63 | Supported | Not on Pareto frontier |
| GLM-5.1 | 67.0 | $4.4 / 1M output price | 15.24 | Supported | Not on Pareto frontier |
| o3-mini | 46.9 | $4.4 / 1M output price | 10.67 | Supported | Not on Pareto frontier |
| Inkling | 67.0 | $4.68 / 1M output price | 14.32 | Supported | Not on Pareto frontier |
| Grok 4.5 | 75.7 | $6 / 1M output price | 12.61 | Supported | Pareto frontier |
| Gemini 3.6 Flash | 75.7 | $7.5 / 1M output price | 10.09 | Supported | Pareto frontier |
| GPT-4.1 | 51.1 | $8 / 1M output price | 6.38 | Supported | Not on Pareto frontier |
| o3 | 47.2 | $8 / 1M output price | 5.9 | Supported | Not on Pareto frontier |
| Gemini 2.5 Pro | 56.9 | $10 / 1M output price | 5.69 | Supported | Not on Pareto frontier |
| GPT-4o | 41.1 | $10 / 1M output price | 4.11 | Supported | Not on Pareto frontier |
| Gemini 3 Pro | 67.6 | $12 / 1M output price | 5.64 | Supported | Not on Pareto frontier |
| GPT-5.3 Codex | 66.2 | $14 / 1M output price | 4.73 | Supported | Not on Pareto frontier |
| GPT-5.2-Codex | 58.4 | $14 / 1M output price | 4.17 | Supported | Not on Pareto frontier |
| Kimi K3 | 80.6 | $15 / 1M output price | 5.37 | Supported | Pareto frontier |
| GPT-5.4 | 73.6 | $15 / 1M output price | 4.9 | Supported | Not on Pareto frontier |
| Claude Sonnet 4.6 | 64.8 | $15 / 1M output price | 4.32 | Supported | Not on Pareto frontier |
| Claude 4 Sonnet | 42.4 | $15 / 1M output price | 2.82 | Supported | Not on Pareto frontier |
| Claude Opus 5 | 83.1 | $25 / 1M output price | 3.32 | Supported | Pareto frontier |
| Claude Opus 4.8 | 76.6 | $25 / 1M output price | 3.06 | Supported | Not on Pareto frontier |
| Claude Opus 4.7 | 72.3 | $25 / 1M output price | 2.89 | Supported | Not on Pareto frontier |
| Claude Opus 4.6 | 68.3 | $25 / 1M output price | 2.73 | Supported | Not on Pareto frontier |
| Claude Opus 4.5 | 63.9 | $25 / 1M output price | 2.56 | Supported | Not on Pareto frontier |
| GPT-5.6 Sol | 82.2 | $30 / 1M output price | 2.74 | Supported | Not on Pareto frontier |
| Claude Mythos 5 | 83.4 | $50 / 1M output price | 1.67 | Supported | Pareto frontier |
| Claude Fable 5 | 83.2 | $50 / 1M output price | 1.66 | Supported | Not on Pareto frontier |
| o1-preview | 48.7 | $60 / 1M output price | 0.81 | Supported | Not on Pareto frontier |
| Claude 4.1 Opus | 45.4 | $75 / 1M output price | 0.6 | Supported | Not on Pareto frontier |
| Claude 3 Opus | 40.7 | $75 / 1M output price | 0.54 | Supported | Not on Pareto frontier |
| o3-pro | 47.6 | $80 / 1M output price | 0.6 | Supported | Not on Pareto frontier |
| GPT-5.2 Pro | 67.3 | $168 / 1M output price | 0.4 | Supported | Not on Pareto frontier |
Step 3.5 Flash
- Score
- 54.6
- Selected cost
- $0.3 / 1M output price
- Score per dollar
- 181.83
- Evidence
- Supported
- Frontier
- Pareto frontier
Qwen3.5 Flash
- Score
- 47.7
- Selected cost
- $0.4 / 1M output price
- Score per dollar
- 119.12
- Evidence
- Supported
- Frontier
- Not on Pareto frontier
DeepSeek V3.2
- Score
- 55.0
- Selected cost
- $0.42 / 1M output price
- Score per dollar
- 130.86
- Evidence
- Supported
- Frontier
- Pareto frontier
Grok 4.1 Fast
- Score
- 50.9
- Selected cost
- $0.5 / 1M output price
- Score per dollar
- 101.76
- Evidence
- Supported
- Frontier
- Not on Pareto frontier
Mercury 2
- Score
- 48.4
- Selected cost
- $0.75 / 1M output price
- Score per dollar
- 64.55
- Evidence
- Supported
- Frontier
- Not on Pareto frontier
Trinity-Large-Thinking
- Score
- 48.2
- Selected cost
- $0.9 / 1M output price
- Score per dollar
- 53.52
- Evidence
- Supported
- Frontier
- Not on Pareto frontier
GLM-4.5-Air
- Score
- 47.3
- Selected cost
- $1.1 / 1M output price
- Score per dollar
- 43
- Evidence
- Supported
- Frontier
- Not on Pareto frontier
DeepSeek V3
- Score
- 44.6
- Selected cost
- $1.1 / 1M output price
- Score per dollar
- 40.52
- Evidence
- Supported
- Frontier
- Not on Pareto frontier
MiniMax M3
- Score
- 68.7
- Selected cost
- $1.2 / 1M output price
- Score per dollar
- 57.28
- Evidence
- Supported
- Frontier
- Pareto frontier
MiniMax M2.7
- Score
- 63.3
- Selected cost
- $1.2 / 1M output price
- Score per dollar
- 52.74
- Evidence
- Supported
- Frontier
- Not on Pareto frontier
MiniMax M2.5
- Score
- 58.9
- Selected cost
- $1.2 / 1M output price
- Score per dollar
- 49.1
- Evidence
- Supported
- Frontier
- Not on Pareto frontier
GPT-5.4 nano
- Score
- 66.8
- Selected cost
- $1.25 / 1M output price
- Score per dollar
- 53.42
- Evidence
- Supported
- Frontier
- Not on Pareto frontier
Inkling-Small
- Score
- 64.0
- Selected cost
- $1.44 / 1M output price
- Score per dollar
- 44.46
- Evidence
- Supported
- Frontier
- Not on Pareto frontier
Gemini 3.1 Flash-Lite
- Score
- 50.9
- Selected cost
- $1.5 / 1M output price
- Score per dollar
- 33.93
- Evidence
- Supported
- Frontier
- Not on Pareto frontier
Mistral Large 3
- Score
- 49.8
- Selected cost
- $1.5 / 1M output price
- Score per dollar
- 33.21
- Evidence
- Supported
- Frontier
- Not on Pareto frontier
GPT-5 mini
- Score
- 43.3
- Selected cost
- $2 / 1M output price
- Score per dollar
- 21.65
- Evidence
- Supported
- Frontier
- Not on Pareto frontier
DeepSeek-R1
- Score
- 51.2
- Selected cost
- $2.19 / 1M output price
- Score per dollar
- 23.39
- Evidence
- Supported
- Frontier
- Not on Pareto frontier
Gemini 3.5 Flash-Lite
- Score
- 65.5
- Selected cost
- $2.5 / 1M output price
- Score per dollar
- 26.2
- Evidence
- Supported
- Frontier
- Not on Pareto frontier
Grok 4.3
- Score
- 64.3
- Selected cost
- $2.5 / 1M output price
- Score per dollar
- 25.7
- Evidence
- Supported
- Frontier
- Not on Pareto frontier
Gemini 2.5 Flash
- Score
- 47.8
- Selected cost
- $2.5 / 1M output price
- Score per dollar
- 19.1
- Evidence
- Supported
- Frontier
- Not on Pareto frontier
Gemini 3 Flash
- Score
- 60.1
- Selected cost
- $3 / 1M output price
- Score per dollar
- 20.03
- Evidence
- Supported
- Frontier
- Not on Pareto frontier
Kimi K2.5
- Score
- 59.1
- Selected cost
- $3 / 1M output price
- Score per dollar
- 19.71
- Evidence
- Supported
- Frontier
- Not on Pareto frontier
GLM-5
- Score
- 65.9
- Selected cost
- $3.2 / 1M output price
- Score per dollar
- 20.59
- Evidence
- Supported
- Frontier
- Not on Pareto frontier
GLM-5-Turbo
- Score
- 66.3
- Selected cost
- $4 / 1M output price
- Score per dollar
- 16.58
- Evidence
- Supported
- Frontier
- Not on Pareto frontier
GLM-5V-Turbo
- Score
- 62.5
- Selected cost
- $4 / 1M output price
- Score per dollar
- 15.63
- Evidence
- Supported
- Frontier
- Not on Pareto frontier
GLM-5.1
- Score
- 67.0
- Selected cost
- $4.4 / 1M output price
- Score per dollar
- 15.24
- Evidence
- Supported
- Frontier
- Not on Pareto frontier
o3-mini
- Score
- 46.9
- Selected cost
- $4.4 / 1M output price
- Score per dollar
- 10.67
- Evidence
- Supported
- Frontier
- Not on Pareto frontier
Inkling
- Score
- 67.0
- Selected cost
- $4.68 / 1M output price
- Score per dollar
- 14.32
- Evidence
- Supported
- Frontier
- Not on Pareto frontier
Grok 4.5
- Score
- 75.7
- Selected cost
- $6 / 1M output price
- Score per dollar
- 12.61
- Evidence
- Supported
- Frontier
- Pareto frontier
Gemini 3.6 Flash
- Score
- 75.7
- Selected cost
- $7.5 / 1M output price
- Score per dollar
- 10.09
- Evidence
- Supported
- Frontier
- Pareto frontier
GPT-4.1
- Score
- 51.1
- Selected cost
- $8 / 1M output price
- Score per dollar
- 6.38
- Evidence
- Supported
- Frontier
- Not on Pareto frontier
o3
- Score
- 47.2
- Selected cost
- $8 / 1M output price
- Score per dollar
- 5.9
- Evidence
- Supported
- Frontier
- Not on Pareto frontier
Gemini 2.5 Pro
- Score
- 56.9
- Selected cost
- $10 / 1M output price
- Score per dollar
- 5.69
- Evidence
- Supported
- Frontier
- Not on Pareto frontier
GPT-4o
- Score
- 41.1
- Selected cost
- $10 / 1M output price
- Score per dollar
- 4.11
- Evidence
- Supported
- Frontier
- Not on Pareto frontier
Gemini 3 Pro
- Score
- 67.6
- Selected cost
- $12 / 1M output price
- Score per dollar
- 5.64
- Evidence
- Supported
- Frontier
- Not on Pareto frontier
GPT-5.3 Codex
- Score
- 66.2
- Selected cost
- $14 / 1M output price
- Score per dollar
- 4.73
- Evidence
- Supported
- Frontier
- Not on Pareto frontier
GPT-5.2-Codex
- Score
- 58.4
- Selected cost
- $14 / 1M output price
- Score per dollar
- 4.17
- Evidence
- Supported
- Frontier
- Not on Pareto frontier
Kimi K3
- Score
- 80.6
- Selected cost
- $15 / 1M output price
- Score per dollar
- 5.37
- Evidence
- Supported
- Frontier
- Pareto frontier
GPT-5.4
- Score
- 73.6
- Selected cost
- $15 / 1M output price
- Score per dollar
- 4.9
- Evidence
- Supported
- Frontier
- Not on Pareto frontier
Claude Sonnet 4.6
- Score
- 64.8
- Selected cost
- $15 / 1M output price
- Score per dollar
- 4.32
- Evidence
- Supported
- Frontier
- Not on Pareto frontier
Claude 4 Sonnet
- Score
- 42.4
- Selected cost
- $15 / 1M output price
- Score per dollar
- 2.82
- Evidence
- Supported
- Frontier
- Not on Pareto frontier
Claude Opus 5
- Score
- 83.1
- Selected cost
- $25 / 1M output price
- Score per dollar
- 3.32
- Evidence
- Supported
- Frontier
- Pareto frontier
Claude Opus 4.8
- Score
- 76.6
- Selected cost
- $25 / 1M output price
- Score per dollar
- 3.06
- Evidence
- Supported
- Frontier
- Not on Pareto frontier
Claude Opus 4.7
- Score
- 72.3
- Selected cost
- $25 / 1M output price
- Score per dollar
- 2.89
- Evidence
- Supported
- Frontier
- Not on Pareto frontier
Claude Opus 4.6
- Score
- 68.3
- Selected cost
- $25 / 1M output price
- Score per dollar
- 2.73
- Evidence
- Supported
- Frontier
- Not on Pareto frontier
Claude Opus 4.5
- Score
- 63.9
- Selected cost
- $25 / 1M output price
- Score per dollar
- 2.56
- Evidence
- Supported
- Frontier
- Not on Pareto frontier
GPT-5.6 Sol
- Score
- 82.2
- Selected cost
- $30 / 1M output price
- Score per dollar
- 2.74
- Evidence
- Supported
- Frontier
- Not on Pareto frontier
Claude Mythos 5
- Score
- 83.4
- Selected cost
- $50 / 1M output price
- Score per dollar
- 1.67
- Evidence
- Supported
- Frontier
- Pareto frontier
Claude Fable 5
- Score
- 83.2
- Selected cost
- $50 / 1M output price
- Score per dollar
- 1.66
- Evidence
- Supported
- Frontier
- Not on Pareto frontier
o1-preview
- Score
- 48.7
- Selected cost
- $60 / 1M output price
- Score per dollar
- 0.81
- Evidence
- Supported
- Frontier
- Not on Pareto frontier
Claude 4.1 Opus
- Score
- 45.4
- Selected cost
- $75 / 1M output price
- Score per dollar
- 0.6
- Evidence
- Supported
- Frontier
- Not on Pareto frontier
Claude 3 Opus
- Score
- 40.7
- Selected cost
- $75 / 1M output price
- Score per dollar
- 0.54
- Evidence
- Supported
- Frontier
- Not on Pareto frontier
o3-pro
- Score
- 47.6
- Selected cost
- $80 / 1M output price
- Score per dollar
- 0.6
- Evidence
- Supported
- Frontier
- Not on Pareto frontier
GPT-5.2 Pro
- Score
- 67.3
- Selected cost
- $168 / 1M output price
- Score per dollar
- 0.4
- Evidence
- Supported
- Frontier
- Not on Pareto frontier
Method and freshness
Scores are source-published benchmark lanes. Missing score or price facts are unavailable and excluded; published zero prices remain visible without a finite score-per-dollar value.
- Revision
- benchmark_3b14dc55ef5307cfe91356d3ea348101
- Published
- 2026-08-30T02:15:58.000Z
- Checked
- 2026-08-30T02:15:58.000Z
- Data from BenchLM.aiUpdated 2026-08-30T02:15:58.000Z
- LiteLLM corroborationUpdated 2026-08-30T02:15:58.000Z
- Arena ratings from LMArenaUpdated 2026-08-30T02:15:58.000Z
- Catalog and pricing data from OpenRouterUpdated 2026-08-24T00:20:54.139Z
