BenchiusPlaygrounds Hub
Explore chat, image, video, voice, speech, document intelligence, and retrieval across leading model families, then connect them in Workflow Studio. Built-in FinOps is a first-class feature in Benchius. View demo.
Put your agents to work. Try Workflow Studio.
A doodle with a mind of its own.
Demos
Meet Benchius, then watch 3 interactive Workflow Studio walkthroughs and explore the full collection.
Pick the right lab for the idea you want to test.
Chat
10 experiencesOpenAI Chat
Prompt, ground, and stream conversational experiences with Azure OpenAI.
Open playgroundAnthropic Chat
Explore Claude reasoning, prompt caching, web search, and Messages API workflows.
Open playgroundGrok Chat
Compare conversational reasoning and assistant behavior with SpaceXAI models.
Open playgroundCommand Chat
Test enterprise-style command, retrieval, and response workflows.
Open playgroundDeepSeek Playground
Explore DeepSeek-V4-Pro chat completions through a Foundry-compatible endpoint.
Open playgroundMAI-Thinking-1
Explore long-context planning and multi-step reasoning with Microsoft's MAI-Thinking-1 model.
Open playgroundMAI-Code-1.1-Flash
Generate, refactor, and debug code using source files and screenshots with MAI-Code-1.1-Flash.
Open playgroundMAI Cyber
Prepare synthetic security reviews for MAI-Cyber-1-Flash. Live inference awaits an approved private API contract.
Open playgroundMeta Muse Spark
Compare Chat Completions and Responses with native Meta or a compatible Foundry deployment.
Open playgroundKimi Chat
Test Kimi multimodal reasoning, thinking controls, and web-grounded conversations.
Open playgroundImage
5 experiencesOpenAI Image
Generate, edit, and iterate on visual concepts with reference images.
Open playgroundFLUX Image
Explore high-fidelity image generation workflows powered by FLUX.
Open playgroundImagine Image
Turn prompts into bold visual directions with SpaceXAI image generation.
Open playgroundMAI Image
Try Microsoft image generation scenarios from the same hub experience.
Open playgroundMeta Muse Image
Generate, edit, and refine images across turns with Meta Muse Image.
Open playgroundVideo
7 experiencesSora Video
Create, edit, extend, and steer video concepts with OpenAI Sora models.
Open playgroundFLUX 3 Video
Generate, animate, and continue video with synchronized audio and draft-to-final workflows.
Open playgroundImagine Video
Prototype short-form motion ideas and video generation concepts.
Open playgroundLTX Video
Generate and animate scenes with Lightricks LTX Video and reproducible inference controls.
Open playgroundKling 3.0
Generate clips with native audio, multi-shot prompts, and character or object references.
Open playgroundRay 3.2
Guide Luma video with timed keyframes, seamless loops, HDR, and EXR export.
Open playgroundH3
Explore MiniMax H3 video, prompt expansion, keyframe images, and target soundtracks.
Open playgroundVoice & Speech
16 experiencesMAI-Transcribe
Transcribe recordings with speaker diarization, word timestamps, keyword hints, and clean or verbatim text.
Open playgroundMAI-Voice
Create expressive speech with locale-specific voices and SSML style controls. Text is synthesized, not sent to a chat model.
Open playgroundGPT-Transcribe
Transcribe a file or microphone recording with vocabulary and multilingual context. Inspect detected languages in the response.
Open playgroundGPT-Live-Transcribe
Read live microphone captions. Audio turns are committed on a client-side interval or with Finish turn; this model does not use server VAD.
Open playgroundGPT-Realtime
Have a low-latency voice conversation, interrupt spoken responses, or send text in the same live session.
Open playgroundGPT-Realtime-Translate
Translate continuous microphone audio into spoken output and captions. This is an interpreter, not a conversational assistant.
Open playgroundGrok Speech to Speech
Talk with Grok over its native Voice API, with voice selection, automatic or manual turns, and optional native search.
Open playgroundGrok Speech to Text
Transcribe a file or microphone recording with word timestamps, speaker labels, key terms, and optional text formatting.
Open playgroundGrok Text to Speech
Synthesize text with expressive voices, inline speech tags, language selection, and downloadable MP3 or WAV audio.
Open playgroundGPT-Audio Text to Speech
Generate spoken responses or read text aloud with delivery instructions, voice selection, streaming, and local subtitle import.
Open playgroundGPT-Audio Audio Understanding
Ask questions about WAV or MP3 recordings, transcribe, translate, and summarize audio with text responses and follow-up turns.
Open playgroundGPT-Audio Speech to Speech
Send a recording with an optional text prompt and receive a spoken answer through audio-enabled Chat Completions.
Open playgroundGrok Voice 3
Explore Grok Voice 3 with reasoning, all native tool types, precise pronunciation, telephony codecs, binary audio, scripted speech, and resumable conversations.
Open playgroundGrok Live Transcribe
Stream microphone captions with interim results, speaker labels, vocabulary hints, text formatting, VAD, and Smart Turn endpointing.
Open playgroundGrok Voice Utilities
Browse built-in voices, manage authorized custom voices, and inspect SIP registration and call-control requests.
Open playgroundGrok Streaming TTS
Stream text into speech over one reusable WebSocket connection, with per-character timestamps, codec controls, and live PCM playback.
Open playgroundRetrieval
2 experiencesCohere Embed
Embed text and images, inspect their semantic neighborhoods, and pass the dataset to Rerank.
Open playgroundCohere Rerank
Retrieve candidates with vector search, rerank source text and captions, and compare result order.
Open playgroundDocument
2 experiencesMistral Doc AI
Analyze documents and extraction workflows with Mistral document AI.
Open playgroundCohere Parse
Parse document images into Markdown or structured content blocks.
Open playgroundTools
2 experiencesBenchmark Playgrounds
Run one prompt across two or three selected playgrounds, compare their responses side by side, and optionally evaluate and score each participant's result.
Open playgroundWorkflow Studio
Connect model and tool steps, transcribe recordings, and synthesize reviewed voice responses with runnable industry samples.
Open playgroundCompare faster
Move between providers and modalities without rebuilding setup or context.
Prototype richer
Combine chat, media, speech, documents, and retrieval in realistic workflows.
Share clearer demos
Open polished playgrounds that make model capabilities easy to explain.