Poe

Poe now supports @opencode An open-source AI coding terminal that pairs with all major models on Poe. One click login, instant access, no extra configuration. Start building now. poe.com/api/applicatio…

Memory is now available Poe can remember your preferences, interests, and what you're working on across chats, so you don't have to repeat yourself. Works with most official bots, including Claude, GPT, and Gemini. Off by default. Turn it on in Settings → Memory.

Check it out on all platforms at poe.com/Grok-4.3

Grok 4.3 is now available on Poe. It’s a great daily driver, not just a model that wins random benchmarks, but one that holds up for real-life use. It’s also economical enough to use every day, with strong performance for the price. After sharing a lot of feedback with the @xai team, we’ve seen a big step up in how this new model handles complex tasks, especially those involving tool use.

Try them on all platforms and in the Poe API, available at: poe.com/GPT-5.5 and poe.com/GPT-5.5-Pro

GPT-5.5 and GPT-5.5 Pro are live on Poe. We’ve been testing it in early access across complex coding, debugging, and reasoning workflows. On selected internal evals, we saw encouraging improvements: - 12% higher task completion rate - 16% fewer retries on complex prompts - 19% better long-context performance

Tip 3: There is a big jump on coding tasks like - SWE-Bench Pro 64.3% - SWE-Bench Verified 87.6% - TerminalBench 69.4% Altogether, it’s a strong signal that coding performance has improved meaningfully in Opus 4.7 compared to Opus 4.6.

Claude Opus 4.7 is now available on Poe. Anthropic's most capable model yet, with major improvements in coding, reasoning, long-context understanding, tool use, vision, and multi-step agent workflows. You can try it in Poe app on all platforms and in the Poe API at poe.com/Claude-Opus-4.7

Tip 2: Opus 4.7 shows improvement on languages including those with less training data like Yoruba, Igbo and Chichewa. For folks who speak these and other languages this model will be meaningfully better.

Tip 1: Opus 4.7 uses a new tokenizer behind the scenes. In our testing, it can use between 1x and 1.5x more tokens when compared to Opus 4.6, so it’s worth planning accordingly.

Under the hood: it has Extended Thinking Mode, optimized MCP support, multi-agent coordination, and a 1M-token context window. It works really well for workflows that need extensive planning, execution, and self-correction across different steps.

What makes Opus 4.7 different: in our early testing it handles long-running tasks with more rigor, follows instructions more precisely, and verifies its own outputs before reporting back. Designed for work you can hand off with less oversight.

Gemini 3.1 Flash TTS is now available on Poe. A fast, high-quality text-to-speech model from Google. A strong fit for voiceovers, audio content, accessibility features, and adding speech to agent workflows. Try it in Poe app on all platforms and in the Poe API poe.com/Gemini-3.1-Fla…

Seedance 2.0 from ByteDance is now live on Poe. Seedance 2.0 delivers high‑quality video generation across text and image inputs — built for cinematic scenes, consistent characters, and detailed motion control. Seedance 2.0 Fast supports the same modes but is optimized for speed, making it ideal for rapid iteration, quick concept testing, and high‑volume creative workflows. You can try it in Poe app on all platforms and in the Poe API at poe.com/Seedance-2 and poe.com/Seedance-2-Fast

GLM‑5.1 by is now live on Poe. Z.ai’s flagship model for agentic engineering — delivering state‑of‑the‑art results on SWE‑Bench Pro and leading performance in repo generation and real‑world terminal tasks. A major step up from GLM‑5, with significantly stronger coding capabilities across the board. Built for full repository generation, complex codebase refactoring, multi‑step terminal workflows, and autonomous, end‑to‑end engineering tasks. Try it in the Poe app on all platforms and in the Poe API at poe.com/GLM-5.1-FW

Wan 2.7 is now live on Poe! Four video generation modes in one model: Text-to-Video, Image-to-Video, Video Edit, and Reference-to-Video. Supports multi-shot generation, audio-driven output, first-and-last-frame control, and seamless video continuation -- from concept to final cut. You can try it in Poe app on all platforms and in the Poe API at poe.com/Wan-2.7

Gemma 4 is now live on Poe. Google's most capable open model: 31B dense parameters, 256K context window, 140+ languages, multimodal inputs, built-in chain-of-thought reasoning, and function calling support. Apache 2.0 licensed. You can try it in Poe app on all platforms and in the Poe API at poe.com/Gemma-4-31B

Qwen3.6 Plus is now live on Poe. Delivers advanced vision‑language performance from Qwen, with clear gains in code‑heavy workflows like agentic and front‑end coding, plus stronger multimodal understanding including improved OCR and precise object localization. Designed for developers who need reliable coding support alongside sharp visual reasoning in one model. You can try it in Poe app on all platforms and in the Poe API at poe.com/Qwen3.6-Plus

Alibaba's Wan 2.7-Image is now live on Poe. Generate a cohesive set of images from a single prompt with consistent characters, style, and context. Bounding box editing lets you refine specific regions without regenerating the whole image. You can try it in Poe app on all platforms and in the Poe API at poe.com/Wan2.7-Image