ProfileCloudCatalogProductsReportsFounderContact Maxwel Catalog AI Models Back Claude Fable 5 Mythos-class flagship for the longest autonomous tasks and deepest analysis Text ~500B Parameter 1000K Context Claude Opus 5 Multi-step agent workflows and deep multi-file analysis Text ~300B Parameter 1000K Context Claude Sonnet 5 Daily driver for code reviews and general tasks Text ~175B Parameter 200K Context DeepSeek V4 Flash Compact MoE variant for high-volume throughput with very long context at minimal cost Text 284B (13B aktiv) Parameter 1000K Context Open Source DeepSeek V4 Flash Vision Experimental vision model of the V4 generation — run locally on Maxwel hardware, images and documents never leave the device Text 284B (13B aktiv) Parameter 1000K Context Open Source DeepSeek V4-Pro Open-weights frontier model with unified reasoning and chat mode Text / Agent 1.6T (49B aktiv) Parameter 1000K Context Open Source DeepSeek V4.1 Flash DeepSeek's new standard model with image understanding — ahead of V4 Pro on most benchmarks per the vendor, at Flash speed with 1M context Text / Agent 763B (8B/16B aktiv) Parameter 1000K Context Open Source ElevenLabs Audio Isolation Separates speech from background noise and music — for clean podcast and interview recordings Audio ElevenLabs Music Songs, beats, and jingles from text prompts, up to 5 minutes Music ElevenLabs Scribe Transcription with the best word accuracy — speaker diarization, 99 languages, per-word timestamps Speech-to-Text ElevenLabs SFX v2 Sound effects and atmospheres from text prompts Sound ElevenLabs TTS v3 Studio quality for audiobooks and voiceovers in 70+ languages Text-to-Speech Face Swap Swaps faces between two images — for mockups and variants, never for people without consent Image Open Source Flux Fill Pro v1 Extends images beyond their borders and fills regions to match the scene (outpainting/inpainting) Image 32B Parameter Open Source GDPR GLM 5.3 Top-tier model for agentic engineering and SWE-Bench tasks Agent 753B (40B aktiv) Parameter 1000K Context Open Source GLM 5.3 Flash Natively multimodal open-source model (MIT) with 1M context at the lowest price — understands images too Text 320B (18B aktiv) Parameter 1000K Context Open Source Google FILM Interpolates frames between two images — closes video loops and smooths transitions Video Open Source GPT 5.6 Frontier model for agentic workflows with native tool orchestration and multimodal inputs Text ~1.8T Parameter 1050K Context Grok 4.6 Real-time access to X data for current events Text / Agent 1000K Context Kimi K3 Agentic orchestration with long context for multi-step workflows Text 2.8T (16 von 896 Experten aktiv) Parameter 1000K Context Open Source Kling 3 Realistic motion and subject consistency across multiple shots Video Nano Banana 2 Flash Fast image generation for iterative workflows Image Nano Banana Pro 2 Highest detail accuracy for complex scenes with many subjects Image Photo Restoration Restores old photos — resolution, colours and scratches in one pass Image Qwen 3.8 27B Open everyday model with image understanding and 262K context — hosted in the EU Text 27B Parameter 262.144K Context Open Source GDPR Qwen 3.8 Flash Next The fastest model in the lineup: multimodal MoE with just 6B active parameters, built for agents, coding and long-running office workflows Text 125B + 51B N-Gram (6B aktiv) Parameter 262.144K Context Open Source GDPR Qwen 3.8 Max Open frontier model with 2.4 trillion parameters for coding, research and long-horizon agents — the only model of its class available with EU data protection Text 2.4T (95B aktiv) Parameter 262.144K Context Open Source GDPR Recraft V4 Logos, icons, and SVG vector graphics with brand-kit consistency Image Roex Mastering Automatic audio mastering with noise removal and EQ Audio Seedance 2.0 Native joint audio-video with phoneme lip-sync in 8 languages Video Topaz Image restoration and upscaling at strong enlargement Image / Video VEED Fabric 1.0 Turns a photo and an audio track into a talking video — the most accurate lip sync in current comparisons Video Whisper Large v3 Turbo Speech-to-text in 99 languages with high robustness Speech-to-Text 1.5B Parameter Open Source GDPR Deutsch English Light Dark Auto