Fetching from the wire…
01
02
03
04
05
06
07
08
Markets2026-08-14 · source-backed
Gemini 3.7 Flash launched at a 50% introductory cut. Writer claimed 52% agent cost reduction via harness engineering. DeepSeek open-sourced its Harness under MIT so the orchestration layer is free. Hoplite (YC S26) launched to run "software factories" for teams that want to tokenmaxx without infra. Freebuff shipped an ad-subsidized $0 coding agent. Model, harness, runtime, distribution, ads. Token spend became the COGS line every layer of the stack is now competing to compress, and the harness and runtime are now as much a pricing decision as the model.
Each link below shares sources, entities, or timing with this story.
GPT-5.6 Luna went to $0.20 input / $1.20 output per million tokens on July 30. That's an 80% cut. Terra dropped 20%. Luna's input now undercuts Gemini 3.1 Flash-Lite ($0.25/$1.50) and sits at one-fifth of Claude Haiku 4.5's $1 input. Simon Willison covered the announcement and...
The Claude Code source leak was the biggest story in developer tools this week. But the most important analysis didn't come from the people picking through feature flags and Easter eggs. It came from Sebastian Raschka, who read the 512,000 lines of leaked TypeScript and reache...
DeepSeek dropped V4 in mid-June as an open-weight model with a 1-million-token context window, priced at $1.74 per million input tokens, posting near-parity with GPT-5.4 on math and Q&A benchmarks (MindStudio). That's the headline number. The architecture underneath is more in...
DeepSeek-V4-Flash-0731 landed July 31 under MIT with a DSpark speculative-decoding module attached. Terminal Bench 2.1: 82.7. Toolathlon-Verified: 70.3. DSBench-FullStack: 68.7. DeepSWE: 54.4. NL2Repo: 54.2. The model card claims it beats DeepSeek-V4-Pro (Preview) "despite its...
DeepSeek released V4 on April 24 and the numbers demand attention. V4-Pro is 1.6 trillion parameters total with 49 billion active, MIT-licensed, native 1M-token context. It scores 80.6% on SWE-bench Verified, putting it within 0.2 points of Claude Opus 4.6. On Terminal-Bench 2...
Xiaomi released MiMo-V2.5-Pro, a 1.02 trillion parameter mixture-of-experts model (42B active) with 1M token context, fully MIT licensed. In benchmarks, it achieves 63.8% success on agentic tasks using 40-60% fewer tokens than Claude Opus 4.6 or GPT-5.4 for comparable results....
MindPattern daily
One email a day at 7 AM. Sources and a take on every story. Unsubscribe anytime.