Highest intelligence
Most capable
57.6pts
Claude Opus 5.5 (Max)
- Claude Opus 5.5 (Max)57.6 pts
- Claude Sonnet 5.5 (Max)56.0 pts
- Muse Wonder Full (Max)55.8 pts
AI MODELS
Intelligence, price, and speedCompare AI models at a glance.
Find the right model for your work and compare the cost.Last updated 2026-10-19 06:00 KST
Intelligence vs. cost and speed
Intelligence vs. cost and speed · The upper left is cheaper and smarter. Select up to 4 models to compare. The dashed line marks the efficiency frontier.
Side by Side
Compare selected models
Models selected from the chart or leaderboard appear here. A yellow dot marks the best value in each row.
3/4 selected
| Metric | Claude Opus 5.5Max · Anthropic | Muse Wonder FullMax · Meta | GPT-6 AstraMax · OpenAI |
|---|---|---|---|
| Performance | |||
| Intelligence index | 57.6 | 55.8 | 52.7 |
| Coding index | - | - | 76.9 |
| Agentic index | - | - | 51.0 |
| Price (USD / 1M tokens) | |||
| Input | $4 | - | $10 |
| Output | $20 | - | $50 |
| Cached input | $0.20 | - | $1 |
| Cost per Intelligence Index task | $5.98 | - | $3.26 |
| Speed | |||
| Output speed | 98 tok/s | - | 69 tok/s |
| Time to first token | 9.2 min | - | 4.8 min |
| Time to complete 500-token response | 9.3 min | - | 4.9 min |
| Details | |||
| Release date | 2026.09.22 | 2026.10.12 | 2026.09.03 |
Best Picks
Top 3 by use case
For models released in the last 12 months, we use each model’s highest intelligence score variant.
At least 75% of peak intelligence
Lowest cost
$0.13 / task
MiMo-V2.6-Pro
- MiMo-V2.6-Pro$0.13
- Claude Haiku 5.5 (Max)$0.21
- GPT-6.1 Sol (Max)$0.72
At least 50% of peak intelligence
Fastest
1009tok/s
K2 Horizon 375B A23B
- K2 Horizon 375B A23B1009 tok/s
- Muse Spark 1.2 (Xhigh)387 tok/s
- Beam 1.0314 tok/s
Leaderboard
Model leaderboard
The filters above apply here too. Select models with the checkboxes to see them in the comparison tray. Select a column heading to sort.
| Select for comparison | # | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | Claude Opus 5.5Max | Anthropic | 57.6 | - | - | $4 | $20 | $5.98 | 98 tok/s | 2026.09.22 | |
| 2 | Claude Sonnet 5.5Max | Anthropic | 56.0 | - | - | $2 | $10 | $5.46 | 137 tok/s | 2026.09.28 | |
| 3 | Muse Wonder FullMax | Meta | 55.8 | - | - | - | - | - | - | 2026.10.12 | |
| 4 | Claude Fable 5.1Max | Anthropic | 53.4 | 81.6 | 57.9 | $10 | $50 | $7.63 | 70 tok/s | 2026.09.01 | |
| 5 | GPT-6 AstraMax | OpenAI | 52.7 | 76.9 | 51.0 | $10 | $50 | $3.26 | 69 tok/s | 2026.09.03 | |
| 6 | Gemini 4 ArgonHigh | 52.6 | - | - | $2 | $10 | $1.99 | - | 2026.09.30 | ||
| 7 | Muse WonderMax | Meta | 52.4 | - | - | $4 | $20 | $4.52 | 181 tok/s | 2026.10.12 | |
| 8 | GPT-6.1 SolMax | OpenAI | 51.8 | - | - | $2 | $10 | $0.72 | 60 tok/s | 2026.09.29 | |
| 9 | Claude Opus 5Max | Anthropic | 50.8 | 78.0 | 56.5 | $5 | $25 | $5.86 | 58 tok/s | 2026.07.24 | |
| 10 | Claude Fable 5Max, Opus 4.8 Fallback | Anthropic | 49.6 | 76.5 | 50.7 | $10 | $50 | $8.75 | 71 tok/s | 2026.06.09 | |
| 11 | Muse Spark 1.3Max | Meta | 48.1 | 75.8 | 55.5 | $1.25 | $4.25 | $1.6 | 139 tok/s | 2026.09.02 | |
| 12 | GPT-6 SolMax | OpenAI | 47.6 | - | - | $2 | $10 | $1.04 | 121 tok/s | 2026.09.22 | |
| 13 | GPT-5.6 SolMax | OpenAI | 47.0 | 77.4 | 50.2 | $4 | $20 | $1.99 | 89 tok/s | 2026.07.09 | |
| 14 | Grok 4.7Xhigh | SpaceXAI | 46.4 | - | - | $2 | $6 | $3.74 | 94 tok/s | 2026.09.21 | |
| 15 | MiMo-V2.6-Pro | Xiaomi | 46.3 | - | - | $0.43 | $0.87 | $0.13 | 50 tok/s | 2026.09.21 | |
| 16 | Qwen3.8 Max0902 | Alibaba | 45.4 | 76.2 | 56.0 | $2 | $6 | $5.41 | 38 tok/s | 2026.09.02 | |
| 17 | GLM-5.3Max | Z AI | 44.8 | 74.8 | 53.1 | $1.4 | $4.4 | $2.01 | 93 tok/s | 2026.08.18 | |
| 18 | Grok 4.6High | SpaceXAI | 44.3 | 76.8 | 53.0 | $2 | $6 | $1.48 | 66 tok/s | 2026.08.12 | |
| 19 | Step 5 Preview | StepFun | 43.7 | - | - | $1 | $2.7 | $1.03 | 93 tok/s | 2026.09.18 | |
| 20 | Kimi K3Max | Kimi | 43.6 | 76.2 | 50.0 | $3 | $15 | $2 | 43 tok/s | 2026.07.16 | |
| 21 | Claude Haiku 5.5Max | Anthropic | 43.4 | - | - | $0.10 | $0.50 | $0.21 | 234 tok/s | 2026.10.07 | |
| 22 | GPT-5.6 TerraMax | OpenAI | 42.1 | 76.7 | 43.2 | $2 | $12 | $1.4 | 140 tok/s | 2026.07.09 | |
| 23 | Claude Opus 4.8Max | Anthropic | 41.8 | 74.3 | 41.9 | $5 | $25 | $4.08 | 65 tok/s | 2026.05.28 | |
| 24 | GLM-5.3-Flash | Z AI | 41.8 | 71.5 | 50.9 | $0.15 | $0.50 | $0.28 | 78 tok/s | 2026.08.26 | |
| 25 | Ling 3.1 Flash | InclusionAI | 41.1 | - | - | $0.30 | $0.90 | $1.00 | 210 tok/s | 2026.10.01 | |
| 26 | Gemini 3.8 FlashHigh | 40.9 | 76.3 | 40.2 | $0.75 | $3.75 | $1.24 | 158 tok/s | 2026.09.02 | ||
| 27 | Claude Opus 4.7Max | Anthropic | 40.7 | 73.6 | 38.6 | $5 | $25 | - | 52 tok/s | 2026.04.16 | |
| 28 | Qwen3.8 2.4T A95B | Alibaba | 39.9 | 71.9 | 50.1 | $2 | $6 | $2.16 | 38 tok/s | 2026.08.12 | |
| 29 | Qwen3.8-Flash-Next | Alibaba | 39.8 | 73.1 | 53.6 | $0.15 | $0.47 | $0.37 | 53 tok/s | 2026.08.26 | |
| 30 | Muse Spark 1.2Xhigh | Meta | 39.6 | 72.2 | 43.2 | $1.25 | $4.25 | $0.97 | 387 tok/s | 2026.08.05 |
1Claude Opus 5.5
Anthropic · Max
57.6
Intelligence
- Output $/1M
- $20
- Cost / task
- $5.98
- Speed
- 98 tok/s
More metrics
- Coding
- -
- Agentic
- -
- Input $/1M
- $4
- Cached input
- $0.20
- TTFT
- 9.2 min
- Release date
- 2026.09.22
2Claude Sonnet 5.5
Anthropic · Max
56.0
Intelligence
- Output $/1M
- $10
- Cost / task
- $5.46
- Speed
- 137 tok/s
More metrics
- Coding
- -
- Agentic
- -
- Input $/1M
- $2
- Cached input
- $0.10
- TTFT
- 6.4 min
- Release date
- 2026.09.28
3Muse Wonder Full
Meta · Max
55.8
Intelligence
- Output $/1M
- -
- Cost / task
- -
- Speed
- -
More metrics
- Coding
- -
- Agentic
- -
- Input $/1M
- -
- Cached input
- -
- TTFT
- -
- Release date
- 2026.10.12
4Claude Fable 5.1
Anthropic · Max
53.4
Intelligence
- Output $/1M
- $50
- Cost / task
- $7.63
- Speed
- 70 tok/s
More metrics
- Coding
- 81.6
- Agentic
- 57.9
- Input $/1M
- $10
- Cached input
- $0.25
- TTFT
- 2.6 min
- Release date
- 2026.09.01
5GPT-6 Astra
OpenAI · Max
52.7
Intelligence
- Output $/1M
- $50
- Cost / task
- $3.26
- Speed
- 69 tok/s
More metrics
- Coding
- 76.9
- Agentic
- 51.0
- Input $/1M
- $10
- Cached input
- $1
- TTFT
- 4.8 min
- Release date
- 2026.09.03
6Gemini 4 Argon
Google · High
52.6
Intelligence
- Output $/1M
- $10
- Cost / task
- $1.99
- Speed
- -
More metrics
- Coding
- -
- Agentic
- -
- Input $/1M
- $2
- Cached input
- $0.10
- TTFT
- -
- Release date
- 2026.09.30
7Muse Wonder
Meta · Max
52.4
Intelligence
- Output $/1M
- $20
- Cost / task
- $4.52
- Speed
- 181 tok/s
More metrics
- Coding
- -
- Agentic
- -
- Input $/1M
- $4
- Cached input
- $0.20
- TTFT
- 33.4 sec
- Release date
- 2026.10.12
8GPT-6.1 Sol
OpenAI · Max
51.8
Intelligence
- Output $/1M
- $10
- Cost / task
- $0.72
- Speed
- 60 tok/s
More metrics
- Coding
- -
- Agentic
- -
- Input $/1M
- $2
- Cached input
- $0.10
- TTFT
- 3.0 min
- Release date
- 2026.09.29
9Claude Opus 5
Anthropic · Max
50.8
Intelligence
- Output $/1M
- $25
- Cost / task
- $5.86
- Speed
- 58 tok/s
More metrics
- Coding
- 78.0
- Agentic
- 56.5
- Input $/1M
- $5
- Cached input
- $0.50
- TTFT
- 52.9 sec
- Release date
- 2026.07.24
10Claude Fable 5
Anthropic · Max, Opus 4.8 Fallback
49.6
Intelligence
- Output $/1M
- $50
- Cost / task
- $8.75
- Speed
- 71 tok/s
More metrics
- Coding
- 76.5
- Agentic
- 50.7
- Input $/1M
- $10
- Cached input
- $1
- TTFT
- 1.5 min
- Release date
- 2026.06.09
11Muse Spark 1.3
Meta · Max
48.1
Intelligence
- Output $/1M
- $4.25
- Cost / task
- $1.6
- Speed
- 139 tok/s
More metrics
- Coding
- 75.8
- Agentic
- 55.5
- Input $/1M
- $1.25
- Cached input
- $0.15
- TTFT
- 53.6 sec
- Release date
- 2026.09.02
12GPT-6 Sol
OpenAI · Max
47.6
Intelligence
- Output $/1M
- $10
- Cost / task
- $1.04
- Speed
- 121 tok/s
More metrics
- Coding
- -
- Agentic
- -
- Input $/1M
- $2
- Cached input
- $0.20
- TTFT
- 1.8 min
- Release date
- 2026.09.22
13GPT-5.6 Sol
OpenAI · Max
47.0
Intelligence
- Output $/1M
- $20
- Cost / task
- $1.99
- Speed
- 89 tok/s
More metrics
- Coding
- 77.4
- Agentic
- 50.2
- Input $/1M
- $4
- Cached input
- $0.40
- TTFT
- 1.9 min
- Release date
- 2026.07.09
14Grok 4.7
SpaceXAI · Xhigh
46.4
Intelligence
- Output $/1M
- $6
- Cost / task
- $3.74
- Speed
- 94 tok/s
More metrics
- Coding
- -
- Agentic
- -
- Input $/1M
- $2
- Cached input
- $0.50
- TTFT
- 47.4 sec
- Release date
- 2026.09.21
15MiMo-V2.6-Pro
Xiaomi
46.3
Intelligence
- Output $/1M
- $0.87
- Cost / task
- $0.13
- Speed
- 50 tok/s
More metrics
- Coding
- -
- Agentic
- -
- Input $/1M
- $0.43
- Cached input
- $0.0036
- TTFT
- 4.0 sec
- Release date
- 2026.09.21
16Qwen3.8 Max
Alibaba · 0902
45.4
Intelligence
- Output $/1M
- $6
- Cost / task
- $5.41
- Speed
- 38 tok/s
More metrics
- Coding
- 76.2
- Agentic
- 56.0
- Input $/1M
- $2
- Cached input
- $0.25
- TTFT
- 3.1 sec
- Release date
- 2026.09.02
17GLM-5.3
Z AI · Max
44.8
Intelligence
- Output $/1M
- $4.4
- Cost / task
- $2.01
- Speed
- 93 tok/s
More metrics
- Coding
- 74.8
- Agentic
- 53.1
- Input $/1M
- $1.4
- Cached input
- $0.26
- TTFT
- 2.5 sec
- Release date
- 2026.08.18
18Grok 4.6
SpaceXAI · High
44.3
Intelligence
- Output $/1M
- $6
- Cost / task
- $1.48
- Speed
- 66 tok/s
More metrics
- Coding
- 76.8
- Agentic
- 53.0
- Input $/1M
- $2
- Cached input
- $0.50
- TTFT
- 46.0 sec
- Release date
- 2026.08.12
19Step 5 Preview
StepFun
43.7
Intelligence
- Output $/1M
- $2.7
- Cost / task
- $1.03
- Speed
- 93 tok/s
More metrics
- Coding
- -
- Agentic
- -
- Input $/1M
- $1
- Cached input
- $0.10
- TTFT
- 2.2 sec
- Release date
- 2026.09.18
20Kimi K3
Kimi · Max
43.6
Intelligence
- Output $/1M
- $15
- Cost / task
- $2
- Speed
- 43 tok/s
More metrics
- Coding
- 76.2
- Agentic
- 50.0
- Input $/1M
- $3
- Cached input
- $0.30
- TTFT
- 3.6 sec
- Release date
- 2026.07.16
21Claude Haiku 5.5
Anthropic · Max
43.4
Intelligence
- Output $/1M
- $0.50
- Cost / task
- $0.21
- Speed
- 234 tok/s
More metrics
- Coding
- -
- Agentic
- -
- Input $/1M
- $0.10
- Cached input
- $0.01
- TTFT
- 5.1 min
- Release date
- 2026.10.07
22GPT-5.6 Terra
OpenAI · Max
42.1
Intelligence
- Output $/1M
- $12
- Cost / task
- $1.4
- Speed
- 140 tok/s
More metrics
- Coding
- 76.7
- Agentic
- 43.2
- Input $/1M
- $2
- Cached input
- $0.20
- TTFT
- 1.6 min
- Release date
- 2026.07.09
23Claude Opus 4.8
Anthropic · Max
41.8
Intelligence
- Output $/1M
- $25
- Cost / task
- $4.08
- Speed
- 65 tok/s
More metrics
- Coding
- 74.3
- Agentic
- 41.9
- Input $/1M
- $5
- Cached input
- $0.50
- TTFT
- 53.7 sec
- Release date
- 2026.05.28
24GLM-5.3-Flash
Z AI
41.8
Intelligence
- Output $/1M
- $0.50
- Cost / task
- $0.28
- Speed
- 78 tok/s
More metrics
- Coding
- 71.5
- Agentic
- 50.9
- Input $/1M
- $0.15
- Cached input
- $0.03
- TTFT
- 2.7 sec
- Release date
- 2026.08.26
25Ling 3.1 Flash
InclusionAI
41.1
Intelligence
- Output $/1M
- $0.90
- Cost / task
- $1.00
- Speed
- 210 tok/s
More metrics
- Coding
- -
- Agentic
- -
- Input $/1M
- $0.30
- Cached input
- $0.06
- TTFT
- 1.8 sec
- Release date
- 2026.10.01
26Gemini 3.8 Flash
Google · High
40.9
Intelligence
- Output $/1M
- $3.75
- Cost / task
- $1.24
- Speed
- 158 tok/s
More metrics
- Coding
- 76.3
- Agentic
- 40.2
- Input $/1M
- $0.75
- Cached input
- $0.075
- TTFT
- 24.0 sec
- Release date
- 2026.09.02
27Claude Opus 4.7
Anthropic · Max
40.7
Intelligence
- Output $/1M
- $25
- Cost / task
- -
- Speed
- 52 tok/s
More metrics
- Coding
- 73.6
- Agentic
- 38.6
- Input $/1M
- $5
- Cached input
- $0.50
- TTFT
- 20.9 sec
- Release date
- 2026.04.16
28Qwen3.8 2.4T A95B
Alibaba
39.9
Intelligence
- Output $/1M
- $6
- Cost / task
- $2.16
- Speed
- 38 tok/s
More metrics
- Coding
- 71.9
- Agentic
- 50.1
- Input $/1M
- $2
- Cached input
- $0.25
- TTFT
- 3.0 sec
- Release date
- 2026.08.12
29Qwen3.8-Flash-Next
Alibaba
39.8
Intelligence
- Output $/1M
- $0.47
- Cost / task
- $0.37
- Speed
- 53 tok/s
More metrics
- Coding
- 73.1
- Agentic
- 53.6
- Input $/1M
- $0.15
- Cached input
- $0.016
- TTFT
- 2.0 sec
- Release date
- 2026.08.26
30Muse Spark 1.2
Meta · Xhigh
39.6
Intelligence
- Output $/1M
- $4.25
- Cost / task
- $0.97
- Speed
- 387 tok/s
More metrics
- Coding
- 72.2
- Agentic
- 43.2
- Input $/1M
- $1.25
- Cached input
- $0.15
- TTFT
- 12.5 sec
- Release date
- 2026.08.05
30 of 239 shown · 694 totalPrices are USD per 1M tokens · speed is the median across providers · - means unavailable
Data source: Artificial Analysis. Updated daily. Scores are based on independent benchmarks and may differ in real-world use depending on your prompts and setup. Cost per task and speed are available for select models; a dash means the value is unavailable.