New: DeepSeek-V3.2-Exp! DeepSeek's new experimental model delivers strong performance for long-context queries. You can try it at poe.com/DeepSeek-V3...., in the Poe app across platforms, and use it in the Poe API.

Now on Poe: Claude Sonnet 4.5 Anthropic’s most advanced model yet is one of the best in the world for complex tasks and excels in real-world coding. (1/2)

New: You can now branch conversations from any message on Poe. Each branch copies the conversation up to that point into a new chat so you can test prompts, actions, or ideas and steer the direction without starting over. (1/2)

New Report: Our Summer 2025 analysis shows how recent Al model releases have impacted the landscape, revealing usage shifts across text, image, video, and audio. Highlights: Growing momentum for Google, OpenAI’s GPT-5 launch, and intensifying global competition in multimedia gen 🧵

Now available on Poe: GPT-5-Codex! This specialized version of GPT-5 is optimized for real-world dev tasks and agentic coding, with strong performance on benchmarks like SWE-bench for complex engineering workflows. (1/2)

You can now share your private bots with other Poe users of your choosing.  To create a shareable link, open your bot’s Settings and change access to “Only people with access link.” This feature keeps your bot unlisted while giving you more control over who has access.

Sonoma Alpha has arrived on Poe! This highly intelligent, stealth model from a frontier US lab features a 2M token context window, image input, parallel tool support, and comes in reasoning (Sky) and non-reasoning (Dusk) versions. (1/2)

New on Poe: Qwen3-Next-80B! Alibaba’s latest model brings highly optimized inference → strong tokens/sec throughput + low latency, while still excelling on math & coding benchmarks. (1/2)

The new Kimi K2 snapshot is here! Kimi-K2-0905 brings a 256k context window (double the previous version), enhanced coding and math skills, and overall performance improvements. (1/2)

Now on Poe: ElevenLabs Music! This state-of-the-art music model lets you create custom tracks from a simple text prompt. You can describe the style, mood, and even the language to generate songs with lyrics or instrumentals. (1/2)

Gemini 2.5 Flash Image is dominating image generation on Poe right now. In 72 hours it has captured 48% of image model messages, dethroning previous leaders Imagen-4-Ultra (27%→10%) and GPT-Image-1 (25%→13%). (1/2)

Introducing the /compare command! We’ve made it easier to see how different models respond to the same prompt in a single chat. After a response, just type /compare and @-mention any bot(s) in any chat on Poe. Here’s an example comparing image editing across leading bots 🧵

New on Poe: Grok Code Fast! This high-performance, cost-efficient model from xAI is optimized for agentic coding applications. It offers visible reasoning traces, strong steerability, and supports a 256k context window. (1/2)

Gemini 2.5 Flash Image is now available on Poe! In our evaluation, this model delivers best-in-class image editing, combines images and styles seamlessly, and maintains top-tier consistency across multiple revisions. (1/3)

DeepSeek-V3.1 is now on Poe! This highly anticipated model builds on top of DeepSeek-V3 and DeepSeek-R1 to support hybrid thinking mode, smarter tool calling, and improved thinking efficiency. (1/2)

ElevenLabs v3 is now on Poe! With new audio tags, multi-speaker support, and options for over 70 languages, this category leading model gives you more control to create audio that feels more real, natural and expressive. (1/2)

New usage data shows the launch of GPT-5 has marked a major upgrade cycle for our users. It climbed to number one in just 48 hours, growing to capture 33% of all subscriber messages on Poe within its first week. 🧵

GPT-5 is here! We’re excited to make OpenAI’s most advanced model ever available on Poe the morning of its release. It aims to set a new standard for speed, accuracy, and reliability. (1/2)

OpenAI's new gpt-oss is now on Poe! gpt-oss-120b is a state-of-the-art open-source model that delivers exceptional performance in complex reasoning, multi-step problem-solving, and analysis. (1/2)

Claude Opus 4.1 is now on Poe! Anthropic’s best model for real-world coding and reasoning offers a clear upgrade in understanding and remembering context for more accurate, helpful conversations on your most complex topics. (1/2)

Introducing the Poe API for developers, with three important features: 1) Access to all models and bots on Poe across all modalities 2) A standard OpenAI-compatible chat completions interface 3) The ability to purchase points beyond subscription limits 🧵

Now on Poe: Qwen3-235B-2507! The latest Qwen3-235B snapshot shines with top-tier benchmarks, improved performance on reasoning and knowledge tasks, and a large 256k-token context window. (1/2)

We’ve upgraded our GPT Image bots following OpenAI’s high input fidelity release to the API. Now you can make precise edits while preserving fine details and character consistency. (1/2)

Grok 4's adoption slope on Poe is one of the steepest we've seen among reasoning models in the days following launch. It only took three days to beat the previous week-one record holder, o3. (1/2)

Now on Poe: Kimi K2! Kimi K2 ranks as a top open-source model (SWE-bench Verified at 65.8%), excelling at coding and solving technical problems. (1/2)

Grok 4 is now on Poe! It sets new performance records across nearly all benchmarks and has taken the lead among the frontier models. (1/2)

Hailuo 02 is now on Poe! This new video model stands out for its ability to handle real-world physics and motion where other leading-edge models struggle. It supports text and image prompts, and generates videos that are 6 seconds long. (1/2)

New: o3-deep-research and o4-mini-deep-research! We're excited to bring OpenAI's deep research models to Poe. Now you can tackle complex research tasks and then switch to other models for writing, coding, and media generation all in one chat. (1/3)

New on Poe: Phoenix 1.0! This image model from Leonardo AI supports prompts up to 1500 characters, features an optional prompt enhancer to optimize complex instructions, and has 25 different style presents, including Macro, Sketch, and 3D Render. (1/2)

New on Poe: Lyria! Google DeepMind’s music-gen model creates high-fidelity instrumental tracks directly from your text prompts. With detailed control over style, instruments, and tempo, you can transform your creative ideas into music right from chat. (1/2)

New: Seedream 3.0 and Seedance 1.0 Lite! Seedream 3.0 creates crisp, vibrant images, while Seedance 1.0 Lite quickly generates stunning videos with realistic motion. On Poe, you can create images with Seedream, then at-mention Seedance to turn them into videos in one chat. (1/2)

With FLUX Kontext Pro and GPT-Image-1, you can make conversational edits to images while preserving the subject. This includes use cases like altering hairstyles, clothes, adding objects, or replacing backgrounds. Here are a few examples comparing the two models available on Poe 🧵

o3-pro is now on Poe! OpenAI’s o3-pro is its smartest model to date, outshining o3 and o1-pro in math, science, and coding. It features a 200k token context window and excels at complex, multi-step reasoning. (1/2)

New: Retro Diffusion Plus! Specializing in high-quality pixel art, this upgraded model offers 11 new styles, improved prompt accuracy, and enhanced image clarity. You can choose any style or customization options using the ‘Additional input’ button below the chat. (1/2)

Now on Poe: Gemini 2.5 Pro just got an upgrade that is crushing benchmarks (SOTA on HLE, Aider, GPQA, and more). Key improvements include enhanced multi-turn conversation abilities, complex reasoning, sophisticated coding, and creative writing. (1/2)

Since launching Kling 2.1 Pro on Poe a few days ago, we’ve been impressed with its ability to transform images into stunning 1080p videos. You can combine it with image models on Poe to generate videos in a single chat like this example using FLUX Kontext. (1/3)

New on Poe: FLUX Kontext! Its multi-turn editing lets you refine images quickly with text prompts using our “Remix” button - allowing you to preserve character consistency while editing and even restyling your images. (1/3)

Now on Poe: Veo 3! Veo 3 sets a new standard as the most advanced video model yet, producing highly realistic visuals and immersive audio from text and image-based prompts. (1/2)

New: Runway Gen-4 Turbo! Runway’s latest video model lets you generate dynamic, high-quality videos with realistic motion from text or images with impressive speed. (1/3)

Now on Poe: Imagen 4! Imagen 4 is the newest text-to-image model from Google DeepMind. It produces realistic images with sharpened clarity and impressive speed. (1/2)

Claude Opus 4 and Sonnet 4 are now available on Poe! Claude Opus 4 is Anthropic’s most advanced model ever, delivering top-tier performance for coding and complex tasks. (1/3)

The world’s fastest inference for Llama 4 Scout is now on Poe! At over 2,600 tokens per second, this bot allows for near-instant interactions. (1/2)

New report: How have 2025’s breakthrough model launches influenced AI usage patterns? Our latest ecosystem report examines demand and share-of-usage based on millions of interactions on Poe across text, image, video, audio, and reasoning models. Key highlights: 🧵

Gemini 2.5 Pro on Poe has been updated to the new I/O edition from Google DeepMind! This upgrade delivers significantly improved coding abilities and performance, especially for tasks related to web development. (1/2)

New on Poe: Kling 2.0! This advanced video model sets new standards for motion quality and image-to-video capabilities. It excels at transforming images you create on Poe, including those generated with GPT-4o, into high-quality videos. (1/2)

New: Gemini 2.5 Flash! Google DeepMind's latest model has hybrid reasoning capabilities, along with a 1m token context window, multimodal input support, and native search, all while prioritizing speed and cost. (1/2)

Now available: Qwen3! This powerful new model from Alibaba combines reasoning and performance to excel at complex tasks. (1/3

GPT image generation is on Poe! It pushes the boundaries of image quality and prompt adherence. Unique to Poe, you can generate many versions at once with the repeat command, send requests while an image generates, and combine it with any other model. (1/3)

Now available: Grok 3 and Grok 3 Mini! Grok 3 is xAI's flagship model with deep expertise across specialized domains. It gives in-depth answers to complex queries, drawing on its extensive knowledge across specialized fields. (1/3)

o3 and o4-mini are now on Poe! OpenAI's smartest models yet - o3 is designed to take on the most complex problems and offers extensive visual analysis capabilities, while o4‑mini combines reasoning and speed to excel at everyday tasks. (1/2)