Claude Haiku 4.5
claude-haiku-4-5
selectable
- Provider
- anthropic
- Price
- ~$0.019 / text call
- Quality
- 3/5
- Latency
- 1.5s
- Tasks
- structured_extraction, copy_generation, narrative_reasoning
Model and budget
Choose model overrides and per-call budget ceilings used by /demo and Studio on this device. API keys stay in server environment variables.
Valid range: $0.001–$25. Request ceilings affect auto selection; server env remains the fallback when unset.
Effective settings
- Model
- Auto (router)
- Budget
- $0.250 server default
- Server default ceiling
- $0.250
- Server model env
- none
Video
Model and audio settings for video segment generation (fal.ai). Changes take effect on the next segment render — already-cached clips are unaffected.
Generate audio
Veo 3.1 generates audio alongside the clip. Disable when using the Hero soundtrack assembly (not yet shipped — default on).
Sound providers
Session defaults for cue authoring and regeneration. Per-cue edits in Sound Studio override these. Resolver and model choices apply when authoring cues or regenerating generate-resolver cues.
Music
SFX
VO
Domain graph
Explore how capability families, cards, task classes, and the current story layer connect — the global view behind the capability inspector.
Open domain graphModel registry
Read-only inventory of the router registry. Text, video, and audio models are selectable in the sections above; GFX entries remain router-only. Prices are registry estimates, not provider invoices.
Text
3Claude Sonnet 4.5
claude-sonnet-4-5
selectable
- Provider
- anthropic
- Price
- ~$0.058 / text call
- Quality
- 4/5
- Latency
- 2.5s
- Tasks
- structured_extraction, copy_generation, narrative_reasoning
Claude Opus 4.1
claude-opus-4-1
selectable
- Provider
- anthropic
- Price
- ~$0.290 / text call
- Quality
- 5/5
- Latency
- 4.0s
- Tasks
- narrative_reasoning
GFX
1GPT Image 1
gpt-image-1
router-only
- Provider
- openai
- Price
- $0.063 / frame
- Quality
- 4/5
- Latency
- 20.0s
- Tasks
- image_generation
Audio
4GPT-4o mini TTS
gpt-4o-mini-tts
selectable
- Provider
- openai
- Price
- $0.015 / clip
- Quality
- 3/5
- Latency
- 5.0s
- Tasks
- audio_generation
- Cue types
- vo
ElevenLabs Music v1
eleven-music-v1
selectable
- Provider
- elevenlabs
- Price
- $0.060 / clip
- Quality
- 4/5
- Latency
- 20.0s
- Tasks
- audio_generation
- Cue types
- music
ElevenLabs SFX v1
eleven-sfx-v1
selectable
- Provider
- elevenlabs
- Price
- $0.120 / clip
- Quality
- 4/5
- Latency
- 6.0s
- Tasks
- audio_generation
- Cue types
- sfx
ElevenLabs TTS v3
eleven-tts-v3
selectable
- Provider
- elevenlabs
- Price
- $0.015 / clip
- Quality
- 2/5
- Latency
- 8.0s
- Tasks
- audio_generation
- Cue types
- vo
Video
2Veo 3.1 (via fal.ai)
fal-ai/veo3.1/image-to-video
selectable
- Provider
- fal
- Price
- unknown
- Quality
- 5/5
- Latency
- 150.0s
- Tasks
- video_generation
Kling 1.6 Pro i2v (via fal.ai)
fal-ai/kling-video/v1.6/pro/image-to-video
selectable
- Provider
- fal
- Price
- unknown
- Quality
- 4/5
- Latency
- 90.0s
- Tasks
- video_generation