New AI model: Qwen3.8-Omni-Flash
What shipped
1 new release on the tracker — what actually changed versus the model each one replaces, in the makers' own numbers.
Qwen3.8-Omni-Flash
Alibaba · Qwen Omni line · lightweight · Sep 18, 2026
Qwen's first omni-modal model built for agents: it takes text, images, audio and video and returns text, with a 1M-token context window (up from 256K) and output tokens priced about 90% below Qwen3.5-Omni-Plus.
Compared with Qwen3.5-Omni-Plus · the model it replaces
| Completes 7 in 10 multimodal agent tasks WildClawBench-MM — agent tasks that combine audio, video and tool use | 71.0 ▲ up from 34.5 for Qwen3.5-Omni-Plus · +36.5 points |
| Costs about 90% less per output token than the model it replaces Price per 1M tokens — what developers pay, input / output | $0.15 / $0.47 ▲ down from $0.40 / $4.80 for Qwen3.5-Omni-Plus · output 90% cheaper |
Sources: Qwen · QwenCloud (model page and pricing) · TechNode · Artificial Analysis (Qwen3.5 Omni Plus)
Coming next
Where the flagship lines stand in their own release rhythm. No dates — nobody can promise those — just the facts.
| Gemini · Google · last: Gemini 3.1 Pro | DUE 214 days since the last release — usual gap is 132 |
| Mistral Large · Mistral · last: Mistral Large 3 | DUE 293 days since the last release — usual gap is 149 |
| Claude Opus · Anthropic · last: Claude Opus 5 | MID-CYCLE 59 of ~70 days into its usual cycle |
| Muse · Meta · last: Muse Spark 1.3 | MID-CYCLE 19 of ~28 days into its usual cycle |
| Grok · xAI · last: Grok 4.6 | MID-CYCLE 40 of ~72 days into its usual cycle |
Every line we track, where it stands and what its last release changed: Next AI Model.
Updated automatically every morning · every number links to its source · one email a week, only when something shipped. No filler.