15 models across 4 providers

Every major model.
One memory layer.

Switch between models mid-conversation — your CLAIV memory follows you everywhere. All models powered by OpenRouter for maximum reliability and access to the latest releases.

Economy · 4–5 ET/msgBalanced · 6–11 ET/msgFlagship · 12–18 ET/msg

OpenAI

GPT-5 Nano

Fastest responses for everyday questions and high-volume tasks.

32k context·Quick answers, drafts
4 ET/1k

GPT-5 Mini

Pro+

A large step up in quality at only twice the token cost of Nano.

64k context·Writing, coding, research
5 ET/1k

GPT-5

Power+

OpenAI's flagship reasoning model — strong across coding, math, and logic.

128k context·Complex reasoning, code review
9 ET/1k

GPT-5.2

Power+

Highly capable OpenAI model. State-of-the-art on reasoning benchmarks.

128k context·Frontier analysis, hard problems
11 ET/1k

GPT-5.4

Builder

Latest OpenAI model with computer use and a 1M context window.

1M context·Agentic tasks, large codebases
13 ET/1k

Anthropic

Claude Haiku 4.5

Pro+

Anthropic's fastest Claude — nuanced tone and excellent instruction following.

100k context·Tone-sensitive writing, structured tasks
6 ET/1k

Claude Sonnet 4.6

Power+

Exceptional writing quality and nuanced reasoning at a balanced cost.

200k context·Long-form writing, complex analysis
12 ET/1k

Claude Opus 4.6

Builder

Anthropic's most powerful model. Best for expert-level analysis and writing.

200k context·Demanding creative and analytical work
18 ET/1k

Google

Gemini 2.5 Flash-Lite

Ultra-fast with a longer context window than GPT-5 Nano — great for document tasks.

64k context·Summarisation, classification
4 ET/1k

Gemini 2.5 Flash

Pro+

A 128k context window at a low token cost — ideal for long document analysis.

128k context·Long documents, research
5 ET/1k

Gemini 2.5 Pro

Power+

Google's flagship with a massive 200k context. Best for large corpora.

200k context·Massive documents, multimodal reasoning
9 ET/1k

Gemini 3.1 Pro

Builder

Next-generation Google model. Benchmark leader across coding and reasoning.

1M context·Entire codebases, research libraries
10 ET/1k

Open Source

Llama 3.3 70B

Builder

Meta's open-source flagship — strong reasoning and coding at a very low token cost.

128k context·Developer workflows, auditable AI
5 ET/1k

DeepSeek R1

Builder

Optimised for chain-of-thought reasoning, math, and science at frontier quality.

64k context·Math, science, coding challenges
5 ET/1k

DeepSeek V3

Builder

Fast, highly capable open model rivaling frontier models at a fraction of the cost.

256k context·General purpose, coding, analysis
4 ET/1k

All models at a glance

Sorted by Echo Token rate — lowest to highest

ModelProviderET RateMin PlanContext
GPT-5 NanoOpenAI4 ET / 1kFree32k
Gemini 2.5 Flash-LiteGoogle4 ET / 1kFree64k
DeepSeek V3Open Source4 ET / 1kBuilder256k
GPT-5 MiniOpenAI5 ET / 1kPro+64k
Gemini 2.5 FlashGoogle5 ET / 1kPro+128k
Llama 3.3 70BOpen Source5 ET / 1kBuilder128k
DeepSeek R1Open Source5 ET / 1kBuilder64k
Claude Haiku 4.5Anthropic6 ET / 1kPro+100k
GPT-5OpenAI9 ET / 1kPower+128k
Gemini 2.5 ProGoogle9 ET / 1kPower+200k
Gemini 3.1 ProGoogle10 ET / 1kBuilder1M
GPT-5.2OpenAI11 ET / 1kPower+128k
Claude Sonnet 4.6Anthropic12 ET / 1kPower+200k
GPT-5.4OpenAI13 ET / 1kBuilder1M
Claude Opus 4.6Anthropic18 ET / 1kBuilder200k

ET rates per 1,000 combined input + output tokens. Online search adds a flat +5 ET per message. Persistent memory is included at no extra cost on all plans.

Try every model free

Start with GPT-5 Nano and Gemini Flash-Lite on the free plan. Upgrade to unlock Claude Opus, GPT-5.4, and more.