The model race is now a latency race
the model race is now a latency race: gpt-6 astra hit amazon bedrock’s ultrafast mode on september 30 with up to 6x faster inference and 300 tokens a second, while deepseek-v4.1-flash activates about 8 billion parameters at prefill and 16 at decode out of a 552-billion backbone. 58 new models shipped in september alone.
the flagship is no longer the point.
Context
AWS announced on Sep 30, 2026 that OpenAI's GPT-6 Astra supports an UltraFast mode on Amazon Bedrock, a premium tier with up to 6x faster inference. The Bedrock model card says Ultrafast prices are six times Standard prices.
Foxl measured a median of 7.1 seconds against 20.9 for Standard, with direct calls streaming past 300 tokens a second.
The Sep 30 date and the up to 6x speedup match AWS. The 300 tokens a second comes from Foxl's own test, not an AWS figure in the excerpts. DeepSeek V4.1 Flash is described as a 552 billion parameter backbone with a one million token context; the 8 billion prefill and 16 billion decode active parameters, and 58 new models in September, were not seen in the sources read, so unsupported here, not refuted.
Related work
- AWS: GPT-6 Astra now supports UltraFast mode on Amazon Bedrock ↗Announcement.
- AWS: GPT-6 Astra model card ↗Pricing.
- Hugging Face: DeepSeek-V4.1-Flash ↗Model card.
Watch next
- Read the DeepSeek V4.1 Flash KV cache paper.
Sources
- AWS What's New, Sep 30, 2026aws.amazon.com
- Foxl, Oct 1, 2026foxl.ai
- Baseten, Sep 11, 2026baseten.co
Provenance
The note above is reproduced unedited from the original post, first published on Threads on 2 October 2026 at 19:04 IST. Sources are the papers and datasets the note draws on.
View the original post ↗Embed this note
More notes
The air is now being asked to keep its own ledger
the air is now being asked to keep its own ledger: ecmwf’s aifs compo becomes the first ai model to forecast atmospheric composition globally every three hours, cleanair simulates 365 days of pm2.5 over china in ten seconds, and a unified framework maps six pollutants at one kilometer across the whole country. the air now files its own composition report.
read the note →The current is now being asked to draw its own map
the current is now being asked to draw its own map: china’s langya 2.0 predicts six ocean phenomena including internal waves and mesoscale eddies, a deep net called wenhai resolves eddies globally with air sea flux formulas built in, and scripps infers surface currents from the way temperature patterns deform in satellite images. the ocean now files its own circulation report.
read the note →The soil is now being asked to report its own carbon
the soil is now being asked to report its own carbon: a nix color sensor paired with generative data augmentation predicts soil organic carbon without a lab, random forest drives 74 percent of soil health mapping studies, and sentinel 2 tracks five year carbon change across france and italy from 922 samples. the dirt now files its own carbon account.
read the note →