Sakana Fugu Ultra reports 73.7% on SWE-Bench Pro, beating both Claude Opus 4.8 and GPT-5.5. There is a structural reason to be skeptical of that number — Fugu routes queries to those exact models internally. Here is the full breakdown.
Z.ai, the creator of GLM models, recently launched GLM 5.2. It’s an open-weight model that most developers hadn’t followed this
Claude Mythos scores 93.9% on SWE-bench Verified and ships with a 1M token context window — but it costs 5x more than Opus 4.8 and is not publicly available yet. Here is what the upgrade actually means for developers, and when it is worth waiting for.
Anthropic released two powerful models in less than two weeks, and they are not competing for the same users. Claude
Claude Opus 4.8 hit 69.2% on SWE-bench Pro at launch, the highest score recorded on that benchmark at time of
When a company closes a $65 billion funding round and ships a new flagship model on the same morning, something
Friday AI is the only desktop-native coworker that:
⎇ Watches your screen to understand your UI and app architecture.
⎇ Learns your workflow from the dev server to the deployment.
⎇ Actually hits ‘Submit’ to push your code and ship features.
Integrate your entire stack and build full-scale applications while you’re still on your first cup of coffee.
Get 25 credits for free upon sign-up!