Google DeepMind publishes Gemini 3.8 Flash model card focused on software engineering and agentic workflows
cedric_chee · x · 2026-09-02
Google DeepMind has published the model card for Gemini 3.8 Flash, an iteration on Gemini 3.7 Flash within the Gemini 3 family.
Key details:
- Performance gains focus on software engineering and agentic knowledge workflows
- Retains customizable effort levels to trade off quality, cost, and latency
- Inputs span text, images, audio, and video with up to a 1M token context window; output capped at 64K tokens
- The card also covers safety performance, known limitations, and mitigations
Related event: Google DeepMind launches Gemini 3.8 Flash and cybersecurity variant(3 posts)→
More from Models
- Google launches Gemini 3.8 Flash and a security-focused Cyber variant — dair_ai · 2026-09-02
- Gemini 3.8 Flash starts rolling out to the Gemini App — GeminiApp · 2026-09-02
- Fable 5.1 vs Opus 5: same-prompt 3D ocean test shows rain-drop detail, at 2x the price — socialwithaayan · 2026-09-02
- User one-shots a custom teleprompter app with Fable 5.1 in under an hour — socialwithaayan · 2026-09-02
- Gemini learns to 'watch' video selectively: 88% fewer tokens, 66% lower cost — mfckr_eth · 2026-09-02
- Fable 5.1 vs Kimi K3 on the same game prompt: 60% pricier but clearly better frontend — socialwithaayan · 2026-09-02