At Swa, we’re not chasing unicorn dreams or VC hype. We built this in the Pacific Northwest, funded by our own 401k’s, because we believe small teams deserve the same AI leverage as big corporations.
That’s why our pricing is simple: one predictable fee for your whole team, fair rates for everyone, and no features locked behind addons. Pay for what you use, not what you don't.
Most AI tools charge $20–$30 per user - and that adds up fast. With Swa, your entire team gets powerful AI automation, custom agents, shared memory, document-aware reasoning, and multi-model orchestration - all starting at the cost of just two ChatGPT Pro licenses!
With Swa, you have access to everything included with your membership:
These are the default services included with your base token usage. Swa automatically selects the most cost-effective, high-performing open-source models for each task.
| Task Type | Backed By (Examples) | Notes | Token Estimate |
|---|---|---|---|
| General Chat & Tasks | Llama 3.3, DeepSeek-V3, Qwen 2, Scout | Fast, context-aware, model-routed | Standard 1:1 usage |
| Document Retrieval | Llama 3.3, DeepSeek-MoE | Uses hybrid RAG with embedded context | Standard 1:1 usage |
| Light Reasoning | DeepSeek-V2, Mistral-Instruct | Efficient and responsive | Standard 1:1 usage |
| Image Generation | Flux1 (auto-selected tier) | Optimized for token efficiency | ~8,000–12,000 tokens per image |
| Video Generation | Flux1 (short-form only) | Entry-level media automation | ~35,000–50,000 tokens per short clip |
| These functions are included in the standard token pool with no premium multipliers. | |||
In addition to Swa’s standard task coverage, you can opt into premium models and agents for advanced capabilities. These models provide deeper reasoning, more nuanced outputs, or richer knowledge access - but they also require more compute.
To keep pricing simple and predictable, we translate these added costs into a clear token multiplier. Instead of hidden fees or unpredictable charges, you’ll always know exactly how usage affects your token pool.
(See the table below for details on each premium model or service.)
| Model | Provider | Best For | Swa Token Multiplier | |
|---|---|---|---|---|
| Input | Output | |||
| Grok 3 Latest | xAI | Cutting-edge reasoning and up-to-date knowledge | TBD | TBD |
| Grok 3 Live Search | xAI | Real-time information retrieval with search | TBD | TBD |
| GPT‑4.1 Mini | OpenAI | Fast, efficient general-purpose tasks | TBD | TBD |
| Sonnet 3.7 | Anthropic | Balanced depth and speed | TBD | TBD |
| DeepSeek R1 | DeepSeek | Lightweight, affordable queries | TBD | TBD |
| Perplexity Sonar Pro | Perplexity | Research-focused, knowledge-rich responses | TBD | TBD |
| Model | Provider | Best For | Swa Token Multiplier | |
|---|---|---|---|---|
| Input | Output | |||
| DALL·E | OpenAI | High-quality image generation | TBD | TBD |
| Flux1 | Stability AI | Efficient, token-optimized image creation | TBD | TBD |
| Model | Provider | Best For | Swa Token Multiplier | |
|---|---|---|---|---|
| Input | Output | |||
| Veo3 | Google DeepMind | Advanced generative video creation | TBD | TBD |
| Flux1 (Short-form Video) | Stability AI | Entry-level short clip generation | TBD | TBD |
| Tool / App | Provider | Purpose | Swa Token Multiplier | |
|---|---|---|---|---|
| Input | Output | |||
| Coming Soon | TBD | Advanced productivity & automation tools | TBD | TBD |
The available models, features, and integrations may change over time as we evaluate performance, cost, and reliability. Swa automatically uses the most cost-effective and capable model for each task unless a premium is explicitly requested.
You're always in control. Budget caps prevent surprise overages, and detailed usage analytics show you where tokens go.
Why Swa is Different
Ready to make AI work for your business - without per-user price creep?
