Augment Code adds Gemini 3.7 Flash and its first xAI model, Grok 4.6
Augment added Gemini 3.7 Flash and Grok 4.6 (its first xAI model) to its per-token rate table; existing model rates and the $100/mo Business plan are unchanged.
Token-rate table did not list Gemini 3.7 Flash or any xAI model (captured 2026-08-11)
Gemini 3.7 Flash added at $0.375/$1.875 per MTok in/out; Grok 4.6 added at $2.00/$6.00 per MTok in/out, cache write $0.00 (captured 2026-08-26)
Augment Code’s published per-token rate table (docs.augmentcode.com/models/token-based-pricing) grew by two models between the 2026-08-11 and 2026-08-26 captures, with no change to any existing model’s rate and no change to the flat $100/month Business plan or custom Enterprise pricing.
Gemini 3.7 Flash joined the roster at $0.375 input / $1.875 output / $0.0375 cache read / $0.375 cache write per million tokens, positioned as a fast, cost-efficient option with a 1M-token context window alongside the existing Gemini 3.1 Pro. Grok 4.6 is the more notable addition: it is the first xAI model Augment has published a rate for, priced at $2.00 input / $6.00 output / $0.50 cache read / $0.00 cache write per million tokens — a $0 cache-write rate that stands out against every other model in the table. Both models draw from the same $100 pooled usage balance and 40% service fee that governs the rest of the catalog.
Business/Enterprise plan prices unchanged ($100/mo flat, up to 50 seats), but the published per-model token table (docs.augmentcode.com/models/token-based-pricing) dropped GPT-5.6 Terra from $2.50/$15.00 to $2.00/$12.00 per million input/output tokens (-20%) and GPT-5.6 Luna from $1.00/$6.00 to $0.20/$1.20 (-80%). Claude Sonnet 5 and Claude Opus 4.5 were added to the roster at existing sibling rates, and the Prism (GPT) route (formerly 'Prism (GPT + Kimi)') dropped Kimi K2.6 in favor of GPT-5.6 Sol/Luna. Augment's illustrative 'typical $100 month' usage mix also shifted from $60 LLM/$24 fee/$16 compute to $70 LLM/$28 fee/$2 compute.