Lab comparison rankings.
A provider-level view of who is strongest across public benchmark coverage, best-category wins, open/proprietary mix, fastest matched models, and value leaders.
Labs covered
25
From public benchmark rows
Top lab score
Anthropic
Avg rank 25.3
Most models
OpenAI
Largest tracked model set
Open-weight depth
Alibaba
9 open models
Provider view
Lab strength, coverage, and portfolio mix.
Average rank is computed across available Arena scores, so labs with fewer public rows should be read with coverage in mind.
Lab leaderboard
| Rank | Lab | Models | Avg rank | Best model | Best value | Fastest |
|---|---|---|---|---|---|---|
| #1 | Anthropic Coverage 40% | 320 open | 25.3Best #1 | Claude Fable 5 | Claude Opus 4.7 | Claude Haiku 4.5 20251001 |
| #2 | Moonshot AI Coverage 52% | 76 open | 41.5Best #1 | Kimi K3 | Kimi K3 | Kimi K2.5 Thinking |
| #3 | MiniMax Coverage 56% | 54 open | 47.1Best #20 | Minimax M3 | Minimax M2.7 | Minimax M2.5 |
| #4 | Z.ai Coverage 54% | 87 open | 48.3Best #4 | GLM 5.2 | GLM 5.2 | GLM 4.7 |
| #5 | Meta Coverage 28% | 40 open | 32.3Best #6 | Muse Spark | Muse Spark 1.1 | Muse Spark 1.1 |
| #6 | Google Coverage 35% | 222 open | 41.4Best #5 | Gemini 3.1 Pro Grounding | Gemini 3.5 Flash High | Gemini 3 Flash |
| #7 | OpenAI Coverage 33% | 420 open | 40.4Best #3 | GPT-5.5 Search | GPT-5.6 Sol xHigh | GPT-5.4 Nano High |
| #8 | Xiaomi Coverage 38% | 53 open | 52.6Best #25 | MiMo V2.5 Pro | MiMo V2.5 Pro | MiMo V2.5 Pro |
| #9 | ByteDance Coverage 17% | 20 open | 27.7Best #14 | Seed 2.1 Pro Preview | No price match | No latency match |
| #10 | Thinky Coverage 33% | 11 open | 48.5Best #38 | Inkling | Inkling | No latency match |
| #11 | xAI Coverage 27% | 150 open | 45.6Best #6 | Grok 4.20 Multi Agent Beta 0309 | Grok 4.5 | Grok 4.3 |
| #12 | Alibaba Coverage 36% | 209 open | 56.5Best #17 | Qwen3.7 Max 20260517 | Qwen3.7 Max Preview | Qwen3 VL 235b A22B Thinking |
| #13 | Perplexity Coverage 11% | 20 open | 27.5Best #26 | Ppl Sonar Pro High | No price match | No latency match |
| #14 | Diffbot Coverage 11% | 11 open | 32Best #32 | Diffbot Small Xl | No price match | No latency match |
| #15 | DeepSeek Coverage 34% | 99 open | 69.4Best #29 | DeepSeek V4 Pro Thinking | DeepSeek V4 Pro | DeepSeek V4 Flash |
| #16 | Baidu Coverage 13% | 50 open | 51.2Best #13 | ERNIE 5.1 | No price match | No latency match |
| #17 | IBM Coverage 44% | 11 open | 94Best #94 | Granite 4.1 8b | Granite 4.1 8b | Granite 4.1 8b |
| #18 | Inception AI Coverage 44% | 10 open | 95Best #95 | Mercury 2 | Mercury 2 | Mercury 2 |
| #19 | Mistral Coverage 26% | 85 open | 84.5Best #60 | Mistral Large 3 | Mistral Large 3 | Mistral Large 3 |
| #20 | Poolside Coverage 11% | 22 open | 73.5Best #69 | Laguna M.1 | No price match | No latency match |
| #21 | Meituan Coverage 11% | 10 open | 74Best #74 | Longcat Flash Chat 2602 Exp | No price match | No latency match |
| #22 | Tencent Coverage 11% | 30 open | 77.3Best #64 | Hunyuan Hy3 Preview | No price match | No latency match |
| #23 | Amazon Coverage 11% | 10 open | 86Best #86 | Amazon Nova Experimental Chat 26 02 10 | No price match | No latency match |
| #24 | Stepfun Coverage 11% | 31 open | 90.7Best #84 | Step 1o Turbo 202506 | No price match | No latency match |
| #25 | Nvidia Coverage 11% | 11 open | 93Best #93 | Nvidia Nemotron 3 Ultra 550b A55B Nvfp4 | No price match | No latency match |
Anthropic
32 models, 0 open
Avg rank
25.3
Coverage
40%
Moonshot AI
7 models, 6 open
Avg rank
41.5
Coverage
52%
MiniMax
5 models, 4 open
Avg rank
47.1
Coverage
56%
Z.ai
8 models, 7 open
Avg rank
48.3
Coverage
54%
Meta
4 models, 0 open
Avg rank
32.3
Coverage
28%
22 models, 2 open
Avg rank
41.4
Coverage
35%
OpenAI
42 models, 0 open
Avg rank
40.4
Coverage
33%
Xiaomi
5 models, 3 open
Avg rank
52.6
Coverage
38%
ByteDance
2 models, 0 open
Avg rank
27.7
Coverage
17%
Thinky
1 models, 1 open
Avg rank
48.5
Coverage
33%
xAI
15 models, 0 open
Avg rank
45.6
Coverage
27%
Alibaba
20 models, 9 open
Avg rank
56.5
Coverage
36%
Perplexity
2 models, 0 open
Avg rank
27.5
Coverage
11%
Diffbot
1 models, 1 open
Avg rank
32
Coverage
11%
DeepSeek
9 models, 9 open
Avg rank
69.4
Coverage
34%
Baidu
5 models, 0 open
Avg rank
51.2
Coverage
13%
IBM
1 models, 1 open
Avg rank
94
Coverage
44%
Inception AI
1 models, 0 open
Avg rank
95
Coverage
44%
Mistral
8 models, 5 open
Avg rank
84.5
Coverage
26%
Poolside
2 models, 2 open
Avg rank
73.5
Coverage
11%
Meituan
1 models, 0 open
Avg rank
74
Coverage
11%
Tencent
3 models, 0 open
Avg rank
77.3
Coverage
11%
Amazon
1 models, 0 open
Avg rank
86
Coverage
11%
Stepfun
3 models, 1 open
Avg rank
90.7
Coverage
11%
Nvidia
1 models, 1 open
Avg rank
93
Coverage
11%
Benchmark guide
What the scores mean.
A quick reading key for provider-level comparisons, coverage, average rank, and open-weight portfolio signals.
How is the lab score ranked?
Labs are ranked from available public Arena performance with a coverage adjustment, so broad benchmark coverage matters. A lab with one excellent score should not automatically outrank a lab with many strong model rows.
What does average rank mean?
Average rank is computed across the Arena rows available for that lab's tracked models. It is useful for comparing portfolio strength, but it should be read alongside model count and coverage.
Why do open-weight counts matter?
Open-weight counts show how much of a lab's tracked portfolio can plausibly be self-hosted or inspected outside a closed API. It is a portfolio signal, not a quality score by itself.
Why can labs with fewer models move around?
Small portfolios are more sensitive to one strong or weak model. The page keeps model count and coverage visible so lab comparisons are not reduced to a single rank number.