Qwen3.8-Flash Goes Free on B.AI

// 55m agoMODEL RELEASE

Qwen3.8-Flash Goes Free on B.AI

B.AI is offering Alibaba’s hosted Qwen3.8-Flash API at 0 Credits, with multimodal input, agent controls, and a 1M-token context window. Chat access is rolling out separately, and the free offer is temporary.

// ANALYSIS

Free access makes Qwen3.8-Flash an unusually compelling model for prototyping long-context agents, though developers should avoid designing around a promotion that may disappear.

–The 1M-token context window can handle large repositories, document sets, screenshots, and video in fewer retrieval steps.
–Function calling, structured output, caching, and hosted tools make it practical for production-style agent experiments.
–The related Qwen3.8-Flash-Next release exposes a 125B-parameter MoE architecture with only 6B active per token, but the hosted Flash endpoint’s parameters and benchmark results are not independently published. Qwen GitHub repository: https://github.com/QwenLM/Qwen3.8-Flash-Next
–Developers should benchmark latency, rate limits, and reliability before migrating workloads from established paid APIs.

// TAGS

qwen3.8-flashllmmoemultimodallong-contexttool-useapi

DISCOVERED

55m ago

2026-08-27

PUBLISHED

1h ago

2026-08-27

RELEVANCE

9/ 10

AUTHOR

thenameissky1

// KEEP READING

More AI developer news from the feed

EXPLORE FULL FEED

RESEARCH40m ago

Automata from Agent Traces turns traces into automata

Researchers compress agent execution traces from 12 datasets into compact finite-state machines with 7–43 states. The structures predict next actions, identify likely failures, and suggest that deployment harnesses shape agent behavior more than the underlying model.

INFRA50m ago

Meta’s Power Crunch Puts Cerebras in Play

Meta’s 2026 AI infrastructure budget is projected at $130–145 billion, but power and data-center capacity—not cash—are becoming the binding constraints. An X analysis argues that Cerebras could provide additive inference capacity without relying on conventional HBM-heavy GPU supply, though no large Meta deal has been publicly confirmed.

UPDATE1h ago

ChatGPT Makes Temporary Chats Personal, Saveable

OpenAI now lets users personalize a Temporary Chat with existing memories, custom instructions, and plugins, then choose to save the conversation to history. The update adds an escape hatch for valuable work without forcing every session into fully persistent mode.