GLM-5.3
glm-5.31M context
$1.40 / $4.40
per M tokens
- Intelligence Index
- 59.5v4.1
- Coding Index
- 74.8
- Agentic Index
- 59.1
- Arena text Elo
- 1,487 ±10#13 · 3,751 votes
- WebDev Arena Elo
- 1,599#8
- Output speed
- not measured
Models
Every model we serve, with independent benchmark results beside our rate, so the model you pick is a decision and not a guess.
The full rate table is on the price page; context and max-output details are in the docs.
Rankings at a glance
Ordered by Intelligence Index. Tap a model for its full card.
Swipe the table sideways for the arena scores and rates.
| Model | Intelligence | Coding | Arena text | Arena rank | Input | Output |
|---|---|---|---|---|---|---|
| Grok 4.6† | 60.9 | 76.8 | 1,461 | #46 | $2.00 | $6.00 |
| Kimi K3 | 59.7 | 76.2 | 1,489 | #10 | $3.00 | $15.00 |
| GLM-5.3 | 59.5 | 74.8 | 1,487 | #13 | $1.40 | $4.40 |
| Qwen3.8 Max | 58.1 | 71.8 | 1,481 | #19 | $2.00 | $6.00 |
| Muse Spark 1.2† | 56.8 | 72.2 | 1,498 | #4 | $0.075 | $0.15 |
| DeepSeek V4 Pro† | 53.2 | 68.8 | 1,459 | #50 | $1.32 | $3.96 |
| GLM-5.2 | 52.6 | 68.8 | 1,470 | #34 | $1.05 | $3.30 |
| GPT 5.6 Luna† | 52.3 | 71.4 | 1,451 | #63 | $0.20 | $1.20 |
| DeepSeek V4 Flash† | 51.8 | 69.1 | 1,436 | #86 | $0.286 | $0.858 |
| Qwen3.7 Max† | 46.7 | 66.0 | 1,474 | #27 | $1.875 | $5.625 |
| MiniMax M3 | 45.4 | 58.6 | 1,443 | #76 | $0.225 | $0.90 |
| Kimi K2.6 | 45.1 | 61.8 | 1,461 | #47 | $0.7125 | $3.00 |
| Kimi K2.7 Code | 43.0 | 60.8 | not ranked | not ranked | $0.7125 | $3.00 |
| MiMo V2.5 Pro | 42.9 | 60.2 | 1,468 | #40 | $0.435 | $0.87 |
| Hy3 | 42.2 | 58.8 | 1,457 | #56 | $0.105 | $0.435 |
| GLM-5.1 | 41.0 | 55.8 | 1,468 | #39 | $1.05 | $3.30 |
| Qwen3.6 Plus | 40.5 | 54.5 | 1,443 | #74 | $0.375 | $2.25 |
| Qwen3.7 Plus | 39.4 | 55.9 | 1,458 | #53 | $0.30 | $1.20 |
| MiniMax M2.7 | 38.9 | 52.6 | 1,416 | #121 | $0.225 | $0.90 |
| MiMo V2.5 | 38.0 | 56.8 | 1,434 | #91 | $0.105 | $0.21 |
| LongCat 2.0 | 34.0 | 45.3 | not ranked | not ranked | $0.225 | $0.90 |
| DeepSeek V4 Flash Vision Exp | not scored | not scored | not ranked | not ranked | $0.44 | $1.32 |
GLM
GLM-5.2
glm-5.21M context
$1.05 / $3.30
per M tokens
- Intelligence Index
- 52.6v4.1
- Coding Index
- 68.8
- Agentic Index
- 45.7
- Arena text Elo
- 1,470 ±5#34 · 30,637 votes
- WebDev Arena Elo
- 1,582#11
- Output speed
- not measured
GLM-5.1
glm-5.1202K context
$1.05 / $3.30
per M tokens
- Intelligence Index
- 41.0v4.1
- Coding Index
- 55.8
- Agentic Index
- 30.6
- Arena text Elo
- 1,468 ±4#39 · 42,267 votes
- WebDev Arena Elo
- 1,509#31
- Output speed
- not measured
Kimi
Kimi K3
kimi-k31M context
$3.00 / $15.00
per M tokens
- Intelligence Index
- 59.7v4.1
- Coding Index
- 76.2
- Agentic Index
- 54.3
- Arena text Elo
- 1,489 ±6#10 · 15,054 votes
- WebDev Arena Elo
- 1,674#2
- Output speed
- 35 tok/sTime to first token 2.9 s
Kimi K2.7 Code
kimi-k2.7-code262K context
$0.7125 / $3.00
per M tokens
- Intelligence Index
- 43.0v4.1
- Coding Index
- 60.8
- Agentic Index
- 30.3
- Arena text Elo
- not ranked
- WebDev Arena Elo
- 1,473#41
- Output speed
- 43 tok/sTime to first token 1.1 s
Kimi K2.6
kimi-k2.6262K context
$0.7125 / $3.00
per M tokens
- Intelligence Index
- 45.1v4.1
- Coding Index
- 61.8
- Agentic Index
- 31.2
- Arena text Elo
- 1,461 ±5#47 · 37,543 votes
- WebDev Arena Elo
- 1,509#32
- Output speed
- not measured
MiMo
MiMo V2.5 Pro
mimo-v2.5-pro1M context
$0.435 / $0.87
per M tokens
- Intelligence Index
- 42.9v4.1
- Coding Index
- 60.2
- Agentic Index
- 29.5
- Arena text Elo
- 1,468 ±4#40 · 54,177 votes
- WebDev Arena Elo
- 1,476#40
- Output speed
- 44 tok/sTime to first token 3.0 s
MiMo V2.5
mimo-v2.51M context
$0.105 / $0.21
per M tokens
- Intelligence Index
- 38.0v4.1
- Coding Index
- 56.8
- Agentic Index
- 24.4
- Arena text Elo
- 1,434 ±4#91 · 44,557 votes
- WebDev Arena Elo
- 1,438#53
- Output speed
- 58 tok/sTime to first token 3.7 s
MiniMax
MiniMax M3
minimax-m31M context
$0.225 / $0.90
per M tokens
- Intelligence Index
- 45.4v4.1
- Coding Index
- 58.6
- Agentic Index
- 36.1
- Arena text Elo
- 1,443 ±4#76 · 41,023 votes
- WebDev Arena Elo
- 1,488#37
- Output speed
- 139 tok/sTime to first token 0.9 s
MiniMax M2.7
minimax-m2.7204K context
$0.225 / $0.90
per M tokens
- Intelligence Index
- 38.9v4.1
- Coding Index
- 52.6
- Agentic Index
- 25.9
- Arena text Elo
- 1,416 ±4#121 · 62,076 votes
- WebDev Arena Elo
- 1,398#67
- Output speed
- not measured
Qwen
Qwen3.8 Max
qwen3.8-max1M context
$2.00 / $6.00
per M tokens
- Intelligence Index
- 58.1v4.1
- Coding Index
- 71.8
- Agentic Index
- 58.4
- Arena text Elo
- 1,481 ±7#19 · 9,955 votes
- WebDev Arena Elo
- 1,669#3
- Output speed
- 23 tok/sTime to first token 1.7 s
Qwen3.7 Plus
qwen3.7-plus1M context
$0.30 / $1.20
per M tokens · up to 200K tokens
- Intelligence Index
- 39.4v4.1
- Coding Index
- 55.9
- Agentic Index
- 20.7
- Arena text Elo
- 1,458 ±5#53 · 34,426 votes
- Coding Arena Elo
- 1,503#57
- Output speed
- 56 tok/sTime to first token 1.3 s
Qwen3.7 Max†
qwen3.7-max1M context · closest matching public build
$1.875 / $5.625
per M tokens
- Intelligence Index
- 46.7v4.1
- Coding Index
- 66.0
- Agentic Index
- 30.9
- Arena text Elo
- 1,474 ±10#27 · 3,712 votes
- WebDev Arena Elo
- 1,517#30
- Output speed
- not measured
Qwen3.6 Plus
qwen3.6-plus1M context
$0.375 / $2.25
per M tokens · up to 200K tokens
- Intelligence Index
- 40.5v4.1
- Coding Index
- 54.5
- Agentic Index
- 29.0
- Arena text Elo
- 1,443 ±4#74 · 45,376 votes
- WebDev Arena Elo
- 1,460#45
- Output speed
- not measured
Hy
Hy3
hy3256K context
$0.105 / $0.435
per M tokens
- Intelligence Index
- 42.2v4.1
- Coding Index
- 58.8
- Agentic Index
- 31.4
- Arena text Elo
- 1,457 ±8#56 · 5,877 votes
- WebDev Arena Elo
- 1,518#28
- Output speed
- 76 tok/sTime to first token 1.8 s
LongCat
LongCat 2.0
longcat-2.01M context
$0.225 / $0.90
per M tokens
- Intelligence Index
- 34.0v4.1
- Coding Index
- 45.3
- Agentic Index
- 22.0
- Arena text Elo
- not ranked
- WebDev Arena Elo
- not ranked
- Output speed
- 41 tok/sTime to first token 1.8 s
Muse
Muse Spark 1.2†
muse-spark-1.21M context · closest matching public build
$0.075 / $0.15
per M tokens
- Intelligence Index
- 56.8v4.1
- Coding Index
- 72.2
- Agentic Index
- 49.3
- Arena text Elo
- 1,498 ±10#4 · 3,257 votes
- WebDev Arena Elo
- 1,534#24
- Output speed
- not measured
DeepSeek
DeepSeek V4 Pro†
deepseek-v4-pro1M context · closest matching public build
$1.32 / $3.96
per M tokens
- Intelligence Index
- 53.2v4.1
- Coding Index
- 68.8
- Agentic Index
- 49.6
- Arena text Elo
- 1,459 ±10#50 · 3,621 votes
- WebDev Arena Elo
- 1,582#12
- Output speed
- 69 tok/sTime to first token 1.3 s
DeepSeek V4 Flash Vision Exp
deepseek-v4-flash-vision-exp1M context · not yet on any leaderboard
$0.44 / $1.32
per M tokens
- Intelligence Index
- not scored
- Coding Index
- not scored
- Agentic Index
- not scored
- Arena text Elo
- not ranked
- WebDev Arena Elo
- not ranked
- Output speed
- not measured
DeepSeek V4 Flash†
deepseek-v4-flash1M context · closest matching public build
$0.286 / $0.858
per M tokens
- Intelligence Index
- 51.8v4.1
- Coding Index
- 69.1
- Agentic Index
- 48.4
- Arena text Elo
- 1,436 ±4#86 · 49,001 votes
- WebDev Arena Elo
- 1,579#13
- Output speed
- 109 tok/sTime to first token 0.9 s
Grok
Grok 4.6†
grok-4.6500K context · closest matching public build
$2.00 / $6.00
per M tokens · up to 200K tokens
- Intelligence Index
- 60.9v4.1
- Coding Index
- 76.8
- Agentic Index
- 58.7
- Arena text Elo
- 1,461 ±10#46 · 3,473 votes
- WebDev Arena Elo
- 1,629#5
- Output speed
- 54 tok/sTime to first token 35.7 s
GPT
GPT 5.6 Luna†
gpt-5.6-luna1M context · closest matching public build
$0.20 / $1.20
per M tokens · up to 200K tokens
- Intelligence Index
- 52.3v4.1
- Coding Index
- 71.4
- Agentic Index
- 46.9
- Arena text Elo
- 1,451 ±5#63 · 20,693 votes
- WebDev Arena Elo
- 1,518#29
- Output speed
- 139 tok/sTime to first token 89.1 s
About these figures
Scores are third-party measurements, refreshed periodically from the sources below, and have no bearing on what we charge: the rates on this page are the rates on the price page, and the meter reads neither leaderboard.
Benchmarks: Artificial Analysis Intelligence Index v4.1, retrieved 2026-08-26 — Source: Artificial Analysis.
Arena scores from the LMArena leaderboard dataset (CC BY 4.0), leaderboard published 2026-08-21 — arena.ai/leaderboard.
† Closest matching public build: the leaderboards score a preview, dated or specific-effort build of this model rather than the exact build we serve, so the figures are for the nearest public entry.