Together cuts Qwen3.8 inference pricing, adds GLM-5.3 to serverless lineup
By Rob Litterst, Co-founder, PricingSaaS
Published
Together’s pricing page · Change detected
What changed
Together AI cut serverless inference pricing for Qwen3.8-2.4T-A95B: input tokens fell from $2.50 to $2.00 per 1M (a 20% cut), output from $6.25 to $6.00 per 1M (4%), and cached input from $0.50 to $0.25 per 1M (50%). Together also added GLM-5.3 to its serverless catalog, priced at $1.40 input / $4.40 output / $0.26 cached input per 1M tokens.
Why it matters
As open-weight releases accelerate, neutral inference platforms face margin pressure from both directions — model labs shipping cheaper, better weights, and hyperscalers undercutting on raw compute. Watch whether cached-input pricing keeps compressing faster than input/output pricing across serverless inference providers generally — that would confirm agentic and long-context workloads, not raw token volume, are becoming the real pricing battleground.
In context
The sharpest number here is the 50% cut to cached-input pricing, not the smaller trims to input/output. Cache pricing is the lever inference platforms use to compete for repeat-context workloads — agents, long chat histories, RAG — without touching headline rates that get quoted in comparisons. Pairing a price cut on an existing model with a same-day new-model launch also shows Together competing on portfolio breadth: keep customers inside the serverless catalog by constantly refreshing which model offers the best price per token, rather than defending any single model's price point.
What people are saying
“Models matching Fable 5 capability at a fraction of the cost is genuinely shifting the economics. We actually dug into the performance:”
View original“The standard defense of premium models is that cheap ones need more attempts. Together AI ran 904 DeepSWE rollouts on GLM-5.3 and Claude Fable 5. Tied on the first try. Then the open model pulls ahead on every retry: 87.6% at four attempts against 84.1%, for $15.96 against $86.52. An open model with the higher ceiling, not just the lower price. Retries were supposed to be what you paid to avoid.”
View original