Newsroom
What changed
Releases, price moves and deprecations, collected once a week by the refresh routine and published as it found them. This is a log, not a feed: nothing is ranked by engagement, nothing is summarised by a model, and every item says where it came from.
A vendor filter is on.Items from other vendors are hidden. Use the filter in the header to show everything.
Sources answering
26/27
Items on record
33
Newsroom checked
2026-10-10
every 3 hours
Prices checked
2026-10-10
weekly
How this is gathered. A routine runs in the cloud every three hours against every vendor feed and changelog inthe source list, and publishes what it found. It asks each source whether anything has changed rather than re-downloading it, so most of those checks cost the vendor a single header exchange — and a check that finds nothing writes nothing, which is why this page is not timestamped eight times a day.
Prices are on a different clock. They are re-read once a week, not every three hours. Vendor pricing pages are heavy and not built for polling, and the figures on them barely move; asking sixteen vendors for a megabyte each, eight times a day, to be told nothing changed is how you get blocked and lose the source. Every price still carries the date it was last confirmed. The rules in full.
Everything collected
33 items.
- Added to the catalogue: gpt-rosalind-research at 5.0 in and 25.0 out per million tokens
- Added to the catalogue: gpt-rosalind-discovery at 5.0 in and 25.0 out per million tokens
- Added to the catalogue: gpt-5.6-cyber at 12.5 in and 75.0 out per million tokens
- Added to the catalogue: Jamba Mini at 0.2 in and 0.4 out per million tokens
- Added to the catalogue: Jamba Large at 2.0 in and 8.0 out per million tokens
- Added to the catalogue: Grok 4.20 (Non-Reasoning) at 1.25 in and 2.5 out per million tokens
- Added to the catalogue: Gemini Omni Flash Preview at 1.5 in and 17.5 out per million tokens
- Added to the catalogue: Gemini 3.1 Flash Live Preview at 0.75 in and 4.5 out per million tokens
- Added to the catalogue: Gemini 3 Flash Preview at 0.5 in and 3.0 out per million tokens
- Added to the catalogue: GPT-5.6 at 4.0 in and 20.0 out per million tokens
- Added to the catalogue: GPT-5.4 Pro at 30.0 in and 180.0 out per million tokens
- Added to the catalogue: GPT-5.3 Codex at 1.75 in and 14.0 out per million tokens
- Added to the catalogue: GPT-5.2 Pro at 21.0 in and 168.0 out per million tokens
- Added to the catalogue: GPT-5 Pro at 15.0 in and 120.0 out per million tokens
- Added to the catalogue: GPT-4o mini at 0.15 in and 0.6 out per million tokens
- Added to the catalogue: GPT-4o at 2.5 in and 10.0 out per million tokens
- Added to the catalogue: GPT-4.1 nano at 0.1 in and 0.4 out per million tokens
- Added to the catalogue: GPT-4.1 mini at 0.4 in and 1.6 out per million tokens
- Added to the catalogue: GPT-4.1 at 2.0 in and 8.0 out per million tokens
- Added to the catalogue: GPT-4 at 30.0 in and 60.0 out per million tokens
- Added to the catalogue: GPT-4 Turbo at 10.0 in and 30.0 out per million tokens
- Added to the catalogue: GPT-3.5-turbo at 0.5 in and 1.5 out per million tokens
- Added to the catalogue: Command R+ at 2.5 in and 10.0 out per million tokens
- Added to the catalogue: Codestral 25.08 at 0.3 in and 0.9 out per million tokens
- Added to the catalogue: Claude Sonnet 4.5 (latest) at 3.0 in and 15.0 out per million tokens
- Gemini 3.7 Flash deprecation:
- Gemini 3.5 Flash deprecation:
- Voice design,
- Gemini 3.8 Flash-Lite TTS
- Gemini 3.8 Flash TTS
- Gemini 2.5 models access update: To ensure reliable performance for
- Gemini 3.8 Live Extended Thinking
- Gemini 3.8 Live (gemini-3.8-live): The default option