Documentation menu
On this page

Worldstream models and pricing

Pick a model for the experience you want to build. Native detail, scene continuity, response speed and billing rules all affect that choice.

Worldstream model comparison

Scroll the table horizontally to compare all model details.

Sequencer model pricing in USD. Third-party routes are planned.
World ModelSequencer priceVideo and audioResponse timingQuality / 5
Sequencer VivaAvailable through Sequencer$40 / active stream hourAbout $0.67 / minute. Viva LLM direction included.768p, 24 fpsGenerated video and audioContinuous streamMeasured startup latency not publishedUnrated
H3 Max Directorfal model. Sequencer route planned.$331.20 / generated hour$0.092 / second. Estimate.480p or 768p, 24 fpsStereo audio5–15s segmentsFirst-frame latency not specifiedUnrated
MiniMax FastH3Reactor model. Sequencer route planned.$144.90 / session hour$0.04025 / second. Estimate.1344 × 768, 24 fpsGenerated audioAbout 1× real-time generationClips queued before playbackUnrated
Visko Orbis StableReactor model. Sequencer route planned.$40.16 / session hour$0.011155 / second. Estimate.832 × 480, 18 fpsUpscaled delivery up to 4K; audioAbout 1.8s chunksPrompt changes at chunk boundariesUnrated
Visko Orbis DynamicReactor model. Sequencer route planned.$86.53 / session hour$0.024035 / second. Estimate.640 × 368, 18 fpsUpscaled delivery up to 4K; audioAbout 1.8s chunksLive delivery resolution changesUnrated

Model setup links lead to Sequencer. Specifications were checked October 13, 2026. The tables show customer-facing Sequencer prices. Viva keeps its existing $40/hour rate.

Third-party routes are planned and their prices are estimates. Live quotes determine the actual rate when a connector becomes available. Hourly figures use 3,600 billable seconds and exclude app infrastructure, taxes and separately priced services. Resolution and frame rate describe output format; they do not establish visual quality.

Read the prices correctly

  • Sequencer: Advertises $40 per active stream hour. The session quote is authoritative. Viva includes built-in LLM world direction. Custom Genesis LLM requests and extra image/audio generation are separately metered. Worldstreams API, Genesis integration.
  • fal H3 Max Director: Bills generated video, with a 60-second minimum. Public sessions currently have a 15-minute ceiling. Director pricing and session limits.
  • Reactor: Bills wall-clock time while a session holds a GPU, including ready-but-idle time and recoverable disconnects. Use the session quote to confirm the applicable rate. Billing rules, current pricing API.

The planned Orbis Stable route is estimated at $40.16 per session hour. Viva is $40 per active stream hour, with built-in LLM direction. The billing clocks differ, so compare the session quote and your usage pattern. A complete experience also needs state storage, user access, LLM logic and audience billing. Read about those responsibilities in the host authority guide.

Read the timing correctly

Startup latency is the time from requesting a session to seeing its first usable frame. Response latency is the time from sending an action to seeing its effect. Generation speed describes how quickly video is produced. Chunk duration describes when a streaming model can accept a new direction.

The timing column reports the measure each provider publishes. It is not a common end-to-end latency benchmark. Queuing, network delivery, buffering and your controller can change the wait a player experiences.

Quality ratings

GENESIS uses a 1–5 rating for each of five criteria. The overall quality score is their mean, shown to one decimal place. Speed and price are reported separately.

Criterion What to check
Visual detail Clear subjects, readable environments and visible artifacts.
Prompt adherence Whether the requested action, setting and constraints appear.
Character consistency Whether identities, clothing and objects remain recognizable.
Scene continuity Whether locations, spatial relationships and previous events carry forward.
Motion and stability Natural movement, temporal consistency and recovery after a direction change.

Use 1 for major failures, 2 for frequent failures, 3 for a usable result, 4 for a consistently strong result and 5 for a reliably excellent result across the test set.

Score at least ten shared scenarios, run each three times for two minutes, and publish the prompts, model/version, resolution, settings, date and recordings. Include the same character, inventory and scene-transition challenges for every model. Have at least two reviewers score recordings without provider labels.

No matched GENESIS benchmark has been completed. Every model is currently unrated, including Sequencer. A provider's demo or a clip-generation leaderboard does not establish its score for a persistent interactive world.

Low-latency LLMs

Viva includes its world-direction LLM. For custom moves, dialogue and structured replies, compare Gemini 3.8 Flash, DeepSeek V4.1 Flash and Claude Haiku 5.5.

Scroll the table horizontally to compare all LLM details.

Selected LLMs. Estimated Sequencer token prices in USD per 1M tokens, input / output.
LLMSequencer priceLatency / generation speedModerated / unmoderated
Viva LLMBuilt-in world directionIncluded with VivaViva stream: $40 / active hourRuns in the stream loopMeasured response latency not publishedProvider safety protectionsHosted world rules also apply
Gemini 3.8 FlashGoogle. Sequencer route planned.$0.8625 / $4.3125Introductory estimate through Dec 31, 2026.Low thinking level availableGENESIS response time not benchmarkedAdjustable safety filtersCore protections remain enabled
DeepSeek V4.1 FlashDeepSeek. Sequencer route planned.$0.345 / $1.38Peak, uncached estimate. Off-peak is half.Non-thinking mode availableGENESIS response time not benchmarkedControls not verifiedCheck the selected route's policy
Claude Haiku 5.5Anthropic. Sequencer route planned.$0.115 / $0.575Prompts up to 100K tokens. Above 100K: $0.575 / $2.875.Built for low latencyGENESIS response time not benchmarkedProvider safety protectionsHost policy also applies

Token rates are estimated Sequencer prices per one million input / output tokens, checked October 13, 2026. Gemini's introductory rate runs through December 31, 2026. DeepSeek shows peak pricing for uncached input; off-peak rates are half. Haiku's prices depend on whether the prompt exceeds 100K tokens. Caching, reasoning and tools can affect the final quote. These external Sequencer routes are planned.

The existing tool.llm.generate service is separately billed, currently $0.01/request by default. It is separate from Viva's included world direction.

Latency and moderation

Gemini 3.8 Flash supports a low thinking level. DeepSeek V4.1 Flash supports non-thinking mode. Claude Haiku 5.5 supports adaptive thinking and is designed for latency-sensitive tasks. Response time depends on the prompt, settings and route. Structured JSON needs to finish and pass validation before GENESIS can apply it. None of these models has a matched GENESIS latency benchmark.

Google exposes adjustable safety filters alongside persistent core protections. DeepSeek's moderation controls have not been verified for this comparison. Your host's content policy also applies.

Connect through Sequencer or bring your own adapter.

Sources and model details

Sequencer routes need tested connectors, quotes and lifecycle metering before they become available. The standalone engine also supports your own adapters. Sequencer model routing, connect your own providers.