Microsoft Cuts Coding Model Price to a Quarter: 138B/5B Sparse MoE with Native Vision Ships to Copilot
Background: The Second Price War in Small Coding Models
On August 11, 2026, Microsoft moved MAI-Code-1.1-Flash into general availability on GitHub Copilot. It succeeds MAI-Code-1-Flash (introduced at Build 2026) as the MAI team's iteration on the small-tier coding model line. According to the official GitHub changelog, the new model improves across coding quality, instruction following, tool use, and performance, while adding native vision support.
Notably, GitHub published a second announcement the same day: MAI-Code-1-Flash enters its deprecation countdown, with retirement set for September 10, 2026. This is not a two-tier parallel offering — it is a direct generational replacement.
Core Details: Parameters, Pricing, and That Token Curve
Per the model card data compiled by LLM Stats, MAI-Code-1.1-Flash specifications are:
- Architecture: Transformer + sparse MoE, 138B total parameters / 5B active
- Context: 256K tokens
- Modalities: Text + image in, text out
- Pretrain cutoff: December 2025; training window March–August 2026
Pricing is the most visible change. GitHub Copilot list rates are $0.20 per million input tokens, $0.02 cached input, and $1.20 output; annual subscribers are billed at a 0.25× premium request multiplier. The official framing is «73% lower than MAI-Code-1-Flash», attributed to combined model-side and serving-side efficiency gains.
The vendor-reported benchmarks (run in the same VS Code / Copilot harness, self-reported) are more interesting:
| Benchmark | 1.1-Flash | Tokens | 1-Flash | Tokens |
|---|---|---|---|---|
| SWE-Bench Verified | 72.6% | 8.6K | 71.6% | 10.8K |
| Terminal Bench 2.1 | 62.9% | 17.0K | 51.7% | 14.2K |
SWE-Bench score rose by only one percentage point, but per-task tokens fell from 10.8K to 8.6K, roughly a 20% reduction. The launch post cites additional figures: +22% on Terminal-Bench 2.1 in Copilot CLI, +15% on .NET tasks, +4% code survival, +9% return visits, with tokens streaming 25% faster and 25% fewer tokens consumed per task.
Vision is the new capability. Internal tests report: Text2WebApp 74.1% pass rate, ScreenShot2WebApp 42.1%, and Vision2Web Level3 at just 11.5%.
Commentary: The Product Being Sold Is Unit Cost, Not Score
Fixating on the 72.6% SWE-Bench figure makes this look like a dull upgrade — one percentage point is close to noise at current leaderboard density. But look at the score-per-token ratio and the story inverts: the same problems, 20% fewer tokens burned, at 73% lower list price. Stacked together, the drop in per-task unit cost is order-of-magnitude.
This tracks the industry's narrative shift through the first half of the year. Competition has slid from «who ranks higher on capability» toward «who is cheaper at equivalent intelligence». Microsoft's stated technical attribution is RL environment scale on the training side (the model card cites over 150,000 RL environments; the launch post says «hundreds of thousands of RL environments in Copilot») plus serving-side optimization — essentially trading post-training engineering volume for inference-time economics.
Adaptive solution-length control deserves particular attention: short outputs for easy asks, more budget for hard ones. It reads like a minor engineering trick, but it targets the core pain point of agentic coding — enormous token waste from over-reasoning on simple tasks.
Several caveats warrant discounting: all benchmarks are vendor self-reported and run in the vendor's own harness, without third-party reproduction. The model card's pricing field reads «to be finalized»; the live Copilot pricing page is authoritative. Business and Enterprise users need an administrator to flip the policy, which is off by default. And Vision2Web Level3 at 11.5% signals that vision-grounded coding remains far from usable in complex scenarios.
So What?
For developers running lightweight coding loops, repo Q&A, refactoring, and tool-using agents inside Copilot day to day, this is a nearly frictionless cost reduction — the 0.25× premium request multiplier in particular means the same quota now covers four times the request volume.
The more interesting signal is structural. When a frontier vendor writes «25% fewer tokens per task» into the headline of a launch post, it indicates that marginal returns on raw capability are yielding to inference economics. The next round of coding model competition may not be about gaining two more points on SWE-Bench, but about halving cost at the same score line.