Z.ai Anthropic DeepSeek Microsoft Head to head on 21 boards: Anthropic places best on 16 , DeepSeek places best on 3 , Z.ai places best on 1 and Microsoft places best on 1 . All 4 ranked on 2 of 26 boards.
Board Read Z.ai Anthropic DeepSeek Microsoft 2 Text Arena Arena rating 1d yesterday Nov 9 1d yesterday #23 1,482 GLM-5.3 #4 1,505 best of the picked Claude Opus 5.5 #32 1,475 DeepSeek V4.1 Flash MultiChallenge Score 1d yesterday Nov 9 1d yesterday #10 67.7% Claude Haiku 5.5 #5 74.9% best of the picked DeepSeek V4P1 Flash 7 MMLU-Pro Accuracy · stale 1d yesterday Nov 9 1d yesterday #29 86.0% GLM-5 #4 89.1% best of the picked Claude Opus 4.6 #33 85.0% DeepSeek-V3.2 #75 76.0% Phi 4 Plus LiveBench Global average 1d yesterday Nov 9 1d yesterday #31 76.1% GLM-5.3 #1 83.4% best of the picked Claude Fable 5.1 #8 81.1% DeepSeek V4.1 Flash AA Intelligence Index 1d yesterday Nov 9 1d yesterday #17 31.7 GLM-5.3 #2 59.1 best of the picked Claude Opus 5.5 #16 32.6 DeepSeek V4.1 Flash GPQA Diamond Accuracy 1d yesterday Nov 9 1d yesterday #28 91.9% GLM-5.2 #3 95.6% best of the picked Claude Sonnet 5.5 #21 93.2% DeepSeek V4.1 Flash Epoch ECI ECI 1d yesterday Nov 9 1d yesterday #29 156.9 GLM-5.3 #3 168.4 best of the picked Claude Opus 5.5 #19 159.3 DeepSeek V4.1 Flash ARC-AGI-2 Score 1d yesterday Nov 9 1d yesterday #28 65.8% GLM 5.3 Flash #3 93.3% best of the picked Claude Opus 5.5 #19 78.1% DeepSeek V4.1 Flash HLE Accuracy 1d yesterday Nov 9 1d yesterday #28 8.3% GLM-4p5 #2 46.5% best of the picked Claude Fable 5.1 4 WebDev Arena Arena rating 1d yesterday Nov 9 1d yesterday #25 1,621 GLM-5.3 #1 1,814 best of the picked Claude Opus 5.5 #14 1,665 DeepSeek V4.1 Flash SWE-bench Verified Resolved · stale 1d yesterday Nov 9 1d yesterday #6 72.8% GLM-5 #1 76.8% best of the picked Claude Opus 4.5 #11 70.0% DeepSeek-V3.2 SWE-Bench Pro Resolved 1d yesterday Nov 9 1d yesterday #11 84.3% GLM-5.3 #1 98.0% best of the picked Claude Opus 5 Aider Polyglot Pass rate · stale 1d yesterday Nov 9 1d yesterday #10 72.0% Claude Opus 4 #8 74.2% best of the picked DeepSeek-V3.2 3 BFCL Overall accuracy · stale 1d yesterday Nov 9 1d yesterday #4 72.4% GLM 4.6 #1 77.5% best of the picked Claude Opus 4.5 #12 56.7% DeepSeek-V3.2 #50 28.8% Phi 4 MCP Atlas Pass rate 1d yesterday Nov 9 1d yesterday #8 84.2% GLM-5.3 #3 87.2% best of the picked Claude Fable 5.1 Search Arena Arena rating 1d yesterday Nov 9 1d yesterday #5 1,229 Claude Fable 5 2 Vision Arena Arena rating 1d yesterday Nov 9 1d yesterday #29 1,272 GLM 5.3 Flash #1 1,309 best of the picked Claude Fable 5 #34 1,263 DeepSeek V4.1 Flash Document Arena Arena rating 1d yesterday Nov 9 1d yesterday #40 1,407 GLM-5v Turbo #1 1,499 best of the picked Claude Fable 5 2 Text-to-Image Arena rating 1d yesterday Nov 9 1d yesterday #66 1,012 GLM Image #6 1,330 best of the picked MAI Image 2.6 Image Edit Arena rating 1d yesterday Nov 9 1d yesterday #4 1,467 MAI Image 2.7 2 Open ASR Average WER 1d yesterday Nov 9 1d yesterday #3 5.55 WER azure-speech-09-2026 (microsoft) Speech Arena Arena rating 1d yesterday Nov 9 1d yesterday #31 1,145 Azure Hd 2.5 1 MTEB Mean task score 1d yesterday Nov 9 1d yesterday #1 74.3% harrier-oss-v1-27b (microsoft) 1 AA Speed Output speed 1d yesterday Nov 9 1d yesterday #5 39 tok/s best of the picked GLM 5.3 Flash #9 31 tok/s Claude Sonnet 5.5 #13 23 tok/s DeepSeek V4 Pro 0813 1 AA Price Blended price 1d yesterday Nov 9 1d yesterday #7 $2.15 GLM-5.3 #1 $0.20 best of the picked Claude Haiku 5.5 #3 $0.53 DeepSeek V4.1 Flash 1 OpenRouter Weekly tokens 1d yesterday Nov 9 1d yesterday #2 13.2T tok GLM 5.3 Flash #6 6.8T tok Claude Opus 5.5 #1 56.0T tok best of the picked DeepSeek V4.1 Flash
Green marks the best placing among the picked where two or more are ranked; ties are marked for each. Ranks are each board's own order (lower is better), scores as the board published them. Dash: not on that board.
Modelsby GlyphZ.ai
Anthropic
DeepSeek
Microsoft