Integrated Providers
OpenAI
Text models (GPT series), image generation (DALL-E 3, GPT Image), video (Sora 2), and speech (TTS, Whisper STT). Text:gpt-5-nano, gpt-5, gpt-5.2, gpt-5.2-proImage:
dall-e-3, gpt-image-1.5TTS:
openai-tts, openai-tts-hd, gpt-4o-mini-ttsSTT:
whisper, gpt-4o-transcribe
Anthropic
Claude model family — long context, nuanced reasoning, and safe outputs. Models:claude-haiku-4.5, claude-sonnet-4.5, claude-opus-4.5
gemini-2.5-flash, gemini-3-pro
MidJourney
The leading model for artistic and stylized image generation. Also supports video. Image:midjourney-v6, midjourney-turboVideo:
mj-video-480p, mj-video-720p
Replicate
Hosts a wide range of open-source models: Flux, Stable Diffusion, Kling, Recraft, MusicGen, and image tools. Image:flux-1.1-pro, flux-kontext-pro, sd-3.5-large, recraft-v3, nano-banana-proVideo:
kling-v2.1, kling-v2.1-pro, minimax-videoMusic:
musicgenTTS:
minimax-ttsTools:
upscale, face-swap, remove-bg, style-transfer, colorize, inpaint, text-ocr
ElevenLabs
Best-in-class voice synthesis, dubbing, and music generation. TTS:elevenlabs-flash, elevenlabs-v2STT:
elevenlabs-scribeMusic:
elevenlabs-music
Luma AI
Video generation (Ray2 series) and image generation (Photon). Image:photon-flash, photon-1Video:
luma-flash2, luma-ray2, luma-ray2-1080p
DeepSeek
Open-weight reasoning models, strong for code and complex problem solving. Models:deepseek-v3, deepseek-r1
xAI (Grok)
Grok model family — real-time knowledge, strong reasoning. Models:grok-3-mini
How Provider Routing Works
When you call/api/v2/generate, NeuralBox routes your request to the correct provider based on the model slug. All requests go through:
- NeuralBox API
- Cloudflare Workers proxy (for geo-compliance)
- Provider API

