GLM-5.3 matches frontier models on DeepSWE at fraction of cost
Together AI's DeepSWE tests show GLM-5.3 ties Claude Fable 5 on pass@1 (69.0% vs 69.7%) and wins pass@4 (87.6% vs 84.1%) at $3.99 per rollout vs $21.63. It trails GPT-5.6 Sol on pass@1 (69.0% vs 72.7%) but leads pass@4 and costs 2.1x less.
How this story unfolded
3 weeks · 4 reports · 10 community posts · 14 of 15 shown
- Aug 3
- Aug 18
- Aug 19
GLM-5.3 hits the API at $1.4/$4.4 per million tokensventurebeat.com
An industrial-scale distillation of models, or subtle benchmaxxing: What developers really think of GLM-5.3thenewstack.io
GLM-5.3 achieves 60 on the Artificial Analysis Intelligence Index, on par with Kimi K3 and up 7 points from GLM-5.2. Once the weights are released it will be tied as the leading open weights model
- Aug 21
- Aug 22
- Aug 23
Daily brief
Get tomorrow's AI brief in your inbox
More stories today
- Perplexity adds granular effort selector to Computer
- Mystery AI model Ox Alpha draws developers with free access
- GMKtec to launch Ryzen AI Max+ PRO 495 hardware at IFA Berlin 2026
- Drew Breunig: Fable's high cost ends the free lunch for coding models
- Grok voice mode adds two new voices: Aurora and Liora