aboutsummaryrefslogtreecommitdiffstats
path: root/docs/superpowers/specs/2026-10-23-cloud-models-design.md
AgeCommit message (Collapse)AuthorFilesLines
2026-10-23cloud: the context size from the provider's /modelsDanilo M.1-2/+5
Hetzner runs vLLM, whose /models reports max_model_len per model (262144); OpenRouter names it context_length. Read when the chat's model changes, as /props is for a local one, so the meter shows a cloud chat's fill and its attachment budget is half of what is free instead of a fixed 24000 characters. Image input already went through: a cloud model's inputs count as unknown. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-23docs: cloud models spec done, both sidesDanilo M.1-1/+1
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-23titles: always from the local model, also in a cloud chatDanilo M.1-1/+3
A cloud model gets no enable_thinking switch and reasoned over the title for about 12 s; the local one answers it with thinking off. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-23docs: reflow the cloud lines, To .md stays localDanilo M.1-0/+2
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-23docs: cloud models, example provider and cloud toolsDanilo M.1-1/+1
Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
2026-10-23docs: spec cloud models, chosen per chatDanilo M.1-0/+152
Providers are OpenAI-compatible [[cloud]] tables with a key_cmd; a chat stores its model; cloud chats get only tools marked cloud = true, and audio is always transcribed locally. Replaces the everything-local rule with no silent cloud fallback. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>