Friday, April 17, 2026

2 stories

4

[AINews] Anthropic Claude Opus 4.7 - literally one step better than 4.6 in every dimension

5d ago · 1 source confirmed single-source

Anthropic launched Claude Opus 4.7, claiming improvements across reasoning, coding, vision, and instruction-following with unchanged pricing.

Anthropic officially released Claude Opus 4.7 as its newest top-tier model, positioning it as superior to Opus 4.6 for long-running tasks, coding, and computer use while maintaining $5/$25 per million token pricing. The model features a new tokenizer, support for images up to 2,576 pixels (~3.75 megapixels), a new xhigh reasoning effort tier, and reported benchmark gains including 64.3% on SWE-Bench Pro (+11 points) and 87.6% on SWE-Bench Verified (+7 points). Independent evaluations showed mixed results on document understanding tasks, with significant improvements on charts but higher costs per page compared to specialized alternatives.

it can accept images up to 2,576 pixels on the long edge (~3.75 megapixels), more than three times as many as prior Claude models

— Latent Space
1

Join us at PyCon US 2026 in Long Beach - we have new AI and security tracks this year

4d ago · 1 source confirmed single-source

PyCon US 2026 in Long Beach will feature new dedicated AI and Security tracks alongside traditional programming.

PyCon US 2026 is scheduled for May 13–19 in Long Beach, California, with a new AI track on Friday and Security track on Saturday. The AI track, chaired by Silona Bonewald and Zac Hatfield-Dodds, covers topics including large language models on laptops, Python async patterns for AI agents, and real-time voice agents. The conference remains community-focused with over 2,000 attendees and includes lightning talks, sprints, and open-space unconference sessions.

It's the least corporate feeling large event I've ever attended.

— Simon Willison