Models and API

One API. All models. All apps.

Text, research, image, music, video, and local inference live behind one ChtSafe gateway. The point is not just model breadth, but protected access, deployment-aware routing, and parity across browser, mobile, desktop, Team Members, Sims, workflows, and API.

What makes this different: ShinrAI-protected access, onion-routed and scrambled request options, encrypted retention, on-prem eligibility, and one OpenAI-compatible API that mirrors what users can do in the apps.
Chat and reasoning Team Members and Sims Workflows and agents Image, music, and video Apps and API parity

ShinrAI protected

Identity and sensitive context are protected before external inference, not only once they are already inside a third-party model boundary.

Onion-routed and scrambled

Endpoint and routing strategy can obscure request origin, improve resilience, and balance privacy, price, and performance.

Encrypted retention

Stored history, files, and workflow context stay inside the same protected retention model instead of becoming plain archived prompts.

On-prem and owned-model path

Selected workloads can move from hosted frontier models toward customer-controlled local inference and enterprise-owned AI assets.

Apps and API parity

The same gateway powers browser, iOS, Android, desktop rollout, workflows, Team Members, Sims, and the OpenAI-compatible API.

Full model catalog

All models in one place.

Use the same protected model surface inside chat, Team Members, Sims, grounded workflows, media studios, and the OpenAI-compatible API. The important part is not just breadth, but that the full stack stays inside one ShinrAI-aware gateway.

Catalog note: this overview is grouped by provider and modality so buyers can see the real product surface clearly. Availability still depends on privacy mode, deployment boundary, and workload type, but everything below sits behind the same protected contract rather than separate public endpoints.

OpenAI

Flagship chat, reasoning, deep research, image, and video families.

Chat and reasoning
GPT-5.4 Enterprise GPT-5.4 Pro GPT-o3 GPT-o3 Pro GPT-4.1
Research flows
OpenAI o3 Deep Research OpenAI o4-mini Deep Research

Also represented in the media stack through GPT Image 1.5, Sora, and Sora Turbo.

Google

Gemini reasoning plus the broader Imagen, Veo, and Gemini Flash image stack.

Fast chat and multimodal work
Gemini 2.5 Flash Gemini 3 Flash
Reasoning and long context
Gemini 2.5 Pro Gemini 3.1 Pro

The same family extends into Imagen, Nano Banana, and Veo 3.1 inside the media categories below.

Anthropic

Claude coverage for fast everyday work up to high-trust long-context reasoning.

Claude lineup
Claude Haiku 4.5 Claude Sonnet 4 Claude Sonnet 4.5 Claude Sonnet 4.6 Claude Opus 4.6

Useful when enterprises want a second frontier family beside OpenAI and Gemini without changing the protection model.

xAI

Grok for fast chat, heavy reasoning, coding, plus image and video extensions.

Chat and reasoning
Grok 4.1 Fast Grok 4.2 Fast Grok 3 Grok 4.2 Grok 4.2 Heavy
Coding and media crossover
Grok Code Fast xAI Grok Aurora Grok Imagine Grok Video

DeepSeek

Cost-efficient chat and reasoning families inside the same protected route policy.

Core models
DeepSeek Chat V3 DeepSeek R1

Useful for balanced cost/performance workloads, internal copilots, and routing strategies that blend price with privacy posture.

Perplexity

Research-native models for web-grounded answers and longer investigation workflows.

Research models
Perplexity Sonar Pro Perplexity Sonar Deep Research Perplexity Sonar Reasoning Pro

These are especially useful when Teams and Sims need live, retrieval-heavy reasoning without leaving the same gateway contract.

Local and customer-owned

Open models, coding specialists, and migration paths toward dedicated hardware or air-gapped deployments.

General and reasoning
Qwen 3 Qwen 3.5 Qwen3-30B-A3B Qwen 3 Next Gemma 3.27B GPT-OSS-20B
Coding and owned-model path
Qwen 3 Coder Qwen 3 Coder MAX Qwen 3 Coder Next Customer fine-tuned workflow models

This is where hosted ChtSafe can evolve into dedicated hardware, customer-owned inference, or fully air-gapped enterprise deployments.

Image models

Image generation, editing, and typography

Protected design drafts, brand assets, text-heavy visuals, edits, and remixes live in one creative stack instead of a separate public toolchain.

7 provider families OpenAI, Google, xAI, Flux, Seedream, Ideogram, and Qwen.
Generate, edit, remix From quick concept frames to higher-fidelity brand visuals and typography.
Apps plus API The same protected image surface is usable in workflows and external products.
OpenAI
GPT Image 1.5
Google / Gemini
Google Fast Google Pro Google Ultra Nano Banana 2 Nano Banana Pro Nano Banana Edit
xAI
xAI Grok Aurora Grok Imagine
FX Black Forest Labs
Flux Flux Max
SD ByteDance / Seedream
Seedream 4.0 Seedream 4.5 Seedream 5 Lite
ID Ideogram
Ideogram V3 Ideogram Remix Ideogram Reframe Ideogram Character
Qwen / Alibaba
Qwen Image Qwen2 Z-Image
Music models

Music generation, extension, lyrics, and post-processing

Music runs through the same protected gateway too, from song generation and lyrics to conversions, stems, and music-video handoff.

Suno V3.5 to V5 Fast generation, longer extensions, and premium higher-control variants.
Lyrics plus extension From quick concept tracks to structured songs with controlled follow-up work.
Post-processing tools Vocals, instrumentals, WAV, MIDI, and music-video output in the same flow.
SU Suno
Suno V3.5 Suno V4 Suno V4.5 Suno V4.5 Plus Suno V4.5 All Suno V5
FX Music tooling on top
Lyrics generation Track extension Add vocals Add instrumental Cover conversion Vocal separation WAV conversion MIDI generation Music video
Video models

Video generation, extension, modification, and cinematic variants

Video sits in the same gateway too, with fast short-form models, cinematic premium providers, and enterprise-friendly protected creative routing.

9 provider families Veo, Runway, Seedance, Kling, Wan, Luma, Sora, Minimax, and Grok Video.
Generate, extend, modify Mix fast social formats with higher-fidelity cinematic and transformation routes.
Protected creative routing Keep prompts, uploads, and retention under the same ShinrAI-aware contract.
Google / Veo
Veo 3.1 Fast Veo 3.1 Quality
RW Runway
Runway Gen-4 Runway Gen-4 Turbo Runway Aleph
SE Seedance
Seedance 1.0 Seedance 1.5 Pro Seedance 2 (Beta)
KL Kling
Kling 3.0 Kling 2.6 Kling 2.1 Kling 2.1 Pro Kling 2.1 Standard Kling 1.6 Pro Kling 1.6 Standard Kling 1.5 Pro
WN Wan
Wan 2.1 480p Wan 2.1 720p Wan 2.2 Turbo Wan 2.5
LU Luma
Luma Ray2 Luma Ray2 Flash
OpenAI / Sora
Sora Sora Turbo
MM Minimax
Hailuo Minimax Video 01
BD ByteDance
CapCut
xAI
Grok Video

Every modality above is available through the same protected platform surface: browser, mobile apps, desktop rollout, workflows, Team Members, Sims, and the OpenAI-compatible API. Privacy mode, routing profile, duration, and resolution differ by provider, but the trust model stays consistent.

OpenAI-compatible gateway

The API mirrors the apps instead of living beside them.

ChtSafe is not only a browser workspace. The same gateway powers mobile apps, desktop rollout, internal tools, and external integrations. That means the “one API” promise also includes media, AI coworkers, secure files, and deployment-aware routing.

  • More than completions: route secure chat, Team Members, Sims, media jobs, and grounded workflows through one protected endpoint.
  • More than one surface: what a user can do in browser, iOS, Android, or desktop can also be integrated into internal systems and products.
  • More than hosted-only: the same API story can move with the deployment boundary as workloads grow toward dedicated hardware or on-prem.
Read API quickstart See app surfaces

* Region pinning is available on request but reduces routing flexibility and part of the default privacy and security envelope.

Representative request

Program the same protected stack your users see in the apps.

POST /v1/chat/completions
model: "gpt-5.4"
messages: [...]
tools: ["grounding", "team_members", "sims", "media"]
surface: "web | ios | android | desktop | api"
deployment_mode: "hosted | dedicated | on-prem"

The point of the gateway is that the contract boundary, trust story, and multimodal model access remain consistent whether work starts in the portal or in code.

Chat and reasoning

Use leading text and reasoning models for protected daily work, secure file interactions, and grounded answers.

Team Members and Sims

Run reusable virtual employees and customer simulations against the same protected model surface rather than maintaining separate AI stacks.

Workflows and agents

Route research, approvals, reporting, and automation jobs through the same model gateway with clear commercial treatment.

Media suite

Image, music, and video are part of the same gateway, pricing model, and deployment story instead of living in a separate creative product.

Review the trust boundary See how ShinrAI, encrypted retention, and deployment modes change what leaves the boundary. See AI coworkers in action Understand how Team Members, Sims, workflows, and agents use the same model gateway. Plan the enterprise model path Move recurring workloads toward dedicated hardware or customer-owned local inference over time.