Tencent Hy4 Brings Vision to AI Gateway

// 2h agoMODEL RELEASE

Tencent Hy4 Brings Vision to AI Gateway

Tencent’s Hy4 is now available as tencent/hy4 through Vercel AI Gateway, adding image input, reasoning, tool calling, structured outputs, computer use, and a 1M-token context window. It is 50% off through December 31, 2026.

// ANALYSIS

Hy4 is compelling as a practical agent backend: vision, long context, and computer use in one API give developers a broad surface to test at unusually low introductory pricing.

–First Hy model with vision for fine-grained image and chart understanding
–1M-token context supports large codebases and document collections
–Tool calling and structured outputs fit coding agents and workflow automation
–AI Gateway adds routing, retries, failover, usage tracking, and cost controls
–Teams should benchmark reliability and task success before the discount expires

// TAGS

tencent-hy4llmmultimodalvisionagenttool-uselong-contextinference

DISCOVERED

2h ago

2026-11-04

PUBLISHED

3h ago

2026-11-04

RELEVANCE

9/ 10

AUTHOR

vercel_dev

// KEEP READING

More AI developer news from the feed

EXPLORE FULL FEED

NEWS49m ago

Hermes Agent Kicks Off $20K Plugin Hackathon

Nous Research has launched a seven-day hackathon challenging developers to extend Hermes Agent through plugins, with $20,000 in cash prizes and additional rewards. Plugins can reshape model routing, tools, integrations, and agent behavior.

UPDATE51m ago

Perplexity Decisions API Adds Fast Web Search

Perplexity is adding Fast Search-backed web grounding to the Decisions API, enabling batch-style classification, routing, and scoring with fresh evidence while returning probabilities instead of prose. Grounding costs $0.001 per search, while Decisions input costs $0.02 per million tokens on the 27B model or $0.01 on Flash.

UPDATE1h ago

Dagger Brings Native CI to Entire

Dagger now integrates natively with Entire, an agent-first Git forge for humans and coding agents. Shared caching and identical checks before and after push aim to keep verification fast as agent-generated changes multiply.