From dfc11780314795179a21f42cc18df87edf1999db Mon Sep 17 00:00:00 2001 From: Pascal Garber Date: Sat, 20 Jun 2026 17:34:09 +0200 Subject: [PATCH] =?UTF-8?q?=F0=9F=AA=99=20feat:=20Add=20Token=20Pricing=20?= =?UTF-8?q?for=20Devstral,=20Voxtral,=20Holo2,=20and=20Mistral=20Medium=20?= =?UTF-8?q?(#13863)?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit These open-weight models are not in tokenValues, so they fall back to defaultRate ($6/1M) for balance/transaction accounting on custom OpenAI-compatible endpoints (e.g. Scaleway, where they are served). Add representative per-1M-token USD rates: - devstral 0.4 / 2.0 (Mistral API pricing) - mistral-medium 1.5 / 7.5 (Mistral API pricing, Medium 3.5) - voxtral(-small) 0.1 / 0.4 (Mistral API pricing, text) - holo2 0.3 / 0.7 (Scaleway Generative APIs public pricing) Generic keys are used so versioned ids (e.g. devstral-2-123b-instruct-2512, mistral-medium-3.5-128b, voxtral-small-24b-2507, holo2-30b-a3b) match via the existing longest-substring lookup. --- packages/data-schemas/src/methods/tx.ts | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/packages/data-schemas/src/methods/tx.ts b/packages/data-schemas/src/methods/tx.ts index 3e77007b9f..a48c7d16a3 100644 --- a/packages/data-schemas/src/methods/tx.ts +++ b/packages/data-schemas/src/methods/tx.ts @@ -217,6 +217,11 @@ export const tokenValues: Record 'grok-4-1-fast': { prompt: 0.2, completion: 0.5 }, 'grok-code-fast': { prompt: 0.2, completion: 1.5 }, codestral: { prompt: 0.3, completion: 0.9 }, + devstral: { prompt: 0.4, completion: 2.0 }, + 'mistral-medium': { prompt: 1.5, completion: 7.5 }, + voxtral: { prompt: 0.1, completion: 0.4 }, + 'voxtral-small': { prompt: 0.1, completion: 0.4 }, + holo2: { prompt: 0.3, completion: 0.7 }, 'ministral-3b': { prompt: 0.04, completion: 0.04 }, 'ministral-8b': { prompt: 0.1, completion: 0.1 }, 'mistral-nemo': { prompt: 0.15, completion: 0.15 },