OpenClaw has added support for Tencent's Hy4 model, the latest release from Tencent's Hunyuan team, bringing vision input, a 1-million-token context window, and reported gains in coding and long-horizon agent tasks to the platform.
OpenClaw · Tencent Hy4 integration
OpenClaw adds Tencent's Hy4 with vision, a 1M-token context window, and three access routes
Hy4, the latest model from Tencent's Hunyuan team, is now usable inside OpenClaw. OpenClaw reports gains in coding and long-horizon agent tasks, but independent benchmarks have not yet been published.
1M
tokens of context in a single session
3
access routes: TokenHub, TokenPlan, OpenRouter
Yes
vision input alongside text
Scale of the 1M-token context window
Each block = 100K tokens. The announcement gives no baseline comparison, so the column shows size only.
10 blocks × 100K tokens. The window can hold very large inputs in one session, such as long codebases, lengthy documents, or extended conversation histories. Long-context reliability has not been independently verified.
How Hy4 reaches OpenClaw users
Tencent TokenHub
Tencent TokenPlan
OpenRouter
→
OpenClaw platform
→
Developers choose Hy4
Pricing, rate limits, and feature differences between the three routes were not specified.
What stands out
Vision input plus a 1M-token window in one model
Reported coding and long-horizon agent gains
Wider distribution through open routing services
More model choice for agentic workflows on OpenClaw
What remains unverified
No independent benchmark results yet
No rival-model comparisons in the announcement
Pricing and rate limits not disclosed
Multi-step agent reliability varies by task
Continue reading The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.
Already purchased? Sign in ✓ Signed in — this article isn’t included in your current plan.Unlocking the full article...