explainx / blog
Cognition launched SWE-1.7 July 8, 2026 — RL-trained on Kimi K2.7 base, 42.3% FrontierCode 1.1, near GPT-5.5 at lower cost. 1000 TPS via Cerebras in Devin. Training recipe: entropy, multi-cluster RL, self-compaction.
Jul 7, 2026
GLM-5.2 launched in June; by July it is the open model developers actually keep using — MIT license on Hugging Face, 1M-token long-horizon coding, and viral praise from tinygrad's George Hotz. Here's the adoption story beyond the export-ban headlines.
Jul 7, 2026
Hy3 is Tencent's answer to the agentic coding arms race: 295B total parameters, only 21B active per token, production-grade tool calling, and open weights under Apache 2.0. Here's what changed since the April preview and where to try it free.
Jun 29, 2026
After HN front-page hype, hands-on tests say Qwen 3.6 27B dense is the local sweet spot — better code than the 35B MoE, runnable at Q8 on 48GB RAM. Full llama.cpp + OpenCode config inside.
On July 8, 2026, Cognition launched SWE-1.7 — the most capable model they have trained, available today in Devin at 1000 tokens per second via Cerebras. The pitch is not raw benchmark crown — it is frontier-class agentic coding at a fraction of the cost, advancing what they call the cost-performance Pareto curve.
The same day Cursor shipped Grok 4.5 and OpenAI rolled out GPT-Live, Cognition's drop targets long-horizon software engineering — the async tasks Devin was built for, not chat-optimized coding assists.
| Question | Answer |
|---|---|
| Launch? | July 8, 2026 in Devin (Web, Desktop, CLI) |
| Speed? | 1000 tok/s via Cerebras |
| Base? | Kimi K2.7 Code + Cognition RL |
| FrontierCode 1.1? | 42.3% (GPT-5.5: 43.0%, Opus 4.8: 46.5%) |
| Terminal-Bench 2.1? | 81.5% (Opus 4.8: 86.9%) |
| SWE-Bench Multilingual? | 77.8% (Opus 4.8: 84.4%) |
| Open weights? | No — Devin platform |
| RL ceiling? | Cognition argues no — big gains after K2.7 RL |
| Benchmark | SWE-1.7 | Kimi K2.7 | GPT-5.5 | Opus 4.8 | Composer 2.5 | GLM-5.2 |
|---|---|---|---|---|---|---|
| FrontierCode 1.1 Main | 42.3% | 30.1% | 43.0% | 46.5% | 25.6% | 24.5% |
| Terminal-Bench 2.1 | 81.5% | 72.7% | 84.2% | 86.9% | 76.0% | 81.0% |
| SWE-Bench Multilingual | 77.8% | 73.5% | 76.8% | 84.4% | 71.6% | 74.5% |
All models evaluated at maximum reasoning effort per Cognition. Terminal-Bench uses Devin CLI for non-Anthropic/OpenAI models; Anthropic via Claude Code, OpenAI via Codex.
SWE-1.7 sits within a few points of GPT-5.5 on FrontierCode while Cognition's cost curve plots it far left on dollars per rollout — the economic story matters for agent products billing per task.
Long RL runs often hit entropy collapse — policies stop exploring, rewards plateau. Cognition uses top-p sampling plus sampling distribution replay so trainer and inference distributions stay aligned.
Only the trainer needs one high-bandwidth cluster. Rollout inference engines run globally; compressed weight deltas (~99% smaller than full weights) sync through object storage in 1–2 minutes for trillion-parameter-class models.
Rollouts reach up to six hours by having the model summarize working state and resume. Alternating phases — unconstrained success-only training vs budget phases — compress verbosity on easy tasks without killing hard-task reasoning.
False positives and reward hacking are treated as poison. Sandboxes are network-restricted; git history stripped; cheating attempts get zero reward regardless of outcome.
Cognition highlights:
| Path | SWE-1.7 / Devin | GLM-5.2 / Kimi OSS |
|---|---|---|
| Access | Devin subscription | API + open weights |
| Speed | 1000 TPS Cerebras | Varies by host |
| Eval focus | FrontierCode merge quality | Public leaderboards |
| Trust | Cognition trustworthiness post | Self-host control |
Teams on Kimi K2.7 in GitHub Copilot get model intelligence without Devin's harness. SWE-1.7 is the bet that harness + RL + inference co-design beats raw weights alone.
SWE-1.7 is Cognition saying RL is not tapped out — not after Kimi K2.7, not after SWE-1.6's 9.4% FrontierCode. The infrastructure story (multi-continent rollouts, entropy control, six-hour compactions) is as important as the scoreboard.
If you live in Devin, you get 1000 TPS frontier-near coding today. If you live in open weights, GLM-5.2 and Kimi still win on control and price — different Pareto point.
Sources: Cognition — SWE-1.7 · Cognition on X (July 8, 2026).
Benchmarks and availability per Cognition's July 8, 2026 post — verify in Devin before procurement decisions.
