Models used
This table summarizes the model and preconfigured setup available through Haloon chat presets.
| Preset name | Underlying model | Reasoning effort | Web access |
|---|---|---|---|
| Chat GPT Instant | openai/gpt-5.4-mini | none | NO |
| Claude | anthropic/claude-sonnet-4.6 | low | YES |
| Claude Fast | anthropic/claude-haiku-4.5 | low | NO |
| Claude Thinking | anthropic/claude-opus-4.7 | high | YES |
| DeepSeek | deepseek/deepseek-v4-pro | none | YES |
| DeepSeek Thinking | deepseek/deepseek-r1-0528 | high | NO |
| Flux 2 | black-forest-labs/flux.2-flex | none | NO |
| Gemini Flash | google/gemini-2.5-flash-lite | none | NO |
| Gemini Pro | google/gemini-3.1-pro-preview | high | NO |
| GPT Codex | openai/gpt-5.3-codex | medium | YES |
| GPT Image 2 | openai/gpt-5.4-image-2 | none | NO |
| GPT Thinking | openai/gpt-5.5 | high | YES |
| Grok Fast | x-ai/grok-4.20 | none | YES |
| Grok Thinking | x-ai/grok-4.3 | high | YES |
| Llama - Fast | meta-llama/llama-4-scout | none | NO |
| Llama - Thinking | meta-llama/llama-4-maverick | high | NO |
| Llama OpenSource | meta-llama/llama-3.3-70b-instruct | none | NO |
| Magistral | mistralai/mistral-large-2512 | high | NO |
| Mistral Instant | mistralai/mistral-medium-3-5 | none | YES |
| Mixtral OpenSource | mistralai/mixtral-8x22b-instruct | none | NO |
| Nano Banana 2 | google/gemini-3.1-flash-image-preview | none | NO |
| Perplexity | perplexity/sonar-pro | none | YES |
| Riverflow | sourceful/riverflow-v2-fast-preview | none | NO |