feat(store,panel,server): price GPT-6 Astra, Fable 5.1 and Mythos 5.1 - #338
Merged
Merged
Conversation
Adds gpt-6-astra, claude-fable-5-1 and claude-mythos-5-1. Sessions on all three were reporting tokens with no cost attached, or with an overstated one. gpt-6-astra bills $10/$50 per MTok with cache reads at $1 and cache writes at $12.50. Prompts over 272K input reprice the entire request at 2x input and 1.5x output, and Fast mode doubles the tariff under the same model id; neither is modelled, because a session total carries no per-request prompt length. Same documented approximation the table already applies to the Gemini context tiers, the Grok 200K surcharge and Opus 5 fast mode. The 5.1 Claude generation is the subtle one. It keeps Fable 5's $10/$50 tariff but cuts the cache read from $1 to $0.25 per MTok, so the prefix fallback onto claude-fable-5 resolves without looking broken while overstating every cache read by 4x. Both ids get their own key, sharing one rate slice so the pair cannot drift; Anthropic left the $0.25 read unconfirmed for Mythos 5.1 at launch, and the shared slice records that assumption in one place. The dot forms claude-fable-5.1 and claude-mythos-5.1 go in the alias table: the prefix fallback requires a "key-" separator, so a dotted minor version can never reach the canonical key on its own. Sources: openai.com/index/gpt-6-astra and OpenAI's model reference page; Anthropic's model pricing table. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_015UmSUq3v73JA3ZJr4nFhm9
displayModel only recognised opus, sonnet and haiku as Claude tier words, so anything else fell through to the title-cased fallback and lost the dot in the version: claude-fable-5-1 rendered as "Fable 5 1". Adding fable and mythos to the tier switch gives "Fable 5.1" and "Mythos 5.1". The 5.0 ids are unaffected — claude-fable-5 and claude-mythos-5 have a single version segment and still render "Fable 5" and "Mythos 5". Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_015UmSUq3v73JA3ZJr4nFhm9
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
What
Prices three newly released models and fixes the panel label for Anthropic's
5.1 generation.
gpt-6-astraclaude-fable-5-1claude-fable-5claude-mythos-5-1claude-mythos-5grok-4.6Grok 4.6 needed no work. Its rates already match xAI's published tariff and it
already has a row in the known-models table.
Why the 5.1 keys matter
The Fable and Mythos fallback resolved, so nothing looked broken. But the 5.1
generation kept the $10/$50 tariff and cut the cache read from $1 to $0.25 per
MTok. Cache reads dominate agent sessions, so every 5.1 session was estimated
about four times too high on the cached portion.
Both ids share one rate slice, following the
gpt56SolPeriodsprecedent, sothe pair cannot drift. Anthropic left the $0.25 read unconfirmed for Mythos 5.1
at launch and the shared slice records that assumption in one place.
The dot forms
claude-fable-5.1andclaude-mythos-5.1go in the alias table.The prefix fallback requires a
key-separator, so a dotted minor version cannever reach the canonical key on its own.
What is not modelled
GPT-6 Astra reprices the entire request at 2x input and 1.5x output above 272K
input tokens, and Fast mode doubles the tariff under the same model id. Neither
is derivable from a session total, so both are noted in the table comment. Same
documented approximation already applied to the Gemini context tiers, the Grok
200K surcharge and Opus 5 fast mode.
Panel label
displayModelonly recognised opus, sonnet and haiku as Claude tier words, soclaude-fable-5-1rendered as "Fable 5 1". Adding fable and mythos to the tierswitch gives "Fable 5.1" and "Mythos 5.1"; the 5.0 ids are unaffected.
Verification
just cipasses end to end: tidy-check, gen-check, vet, lint, race tests,integration tests, build, clean tree.
just qualityand commitlint pass.Against the maintainer's local store,
prosa analytics usage_by_modelnowprices the 25
claude-fable-5-1sessions instead of overstating them.🤖 Generated with Claude Code
https://claude.ai/code/session_015UmSUq3v73JA3ZJr4nFhm9