{
"$type": "site.standard.document",
"bskyPostRef": {
"cid": "bafyreigtui6uyjeybqxqswuckudriigfwu45u5clhqolikoq7zhar5p3wa",
"uri": "at://did:plc:3hmagzdut76xycuvulaezm3j/app.bsky.feed.post/3mkby4xot4fm2"
},
"path": "/2026/04/24/llm-prices.html",
"publishedAt": "2026-04-24T20:42:17.000Z",
"site": "https://www.manton.org",
"textContent": "I’ve been trying to find the right balance of AI models to use in Micro.blog. Currently we use a few different flavors of GPT, including OpenAI’s cheaper “flex” version when speed doesn’t matter much, such as in the background work that powers Inkwell’s Reading Recap feature.\n\nAPI pricing is a little all over the map, so I thought I’d capture some of the recent prices here, for OpenAI and Anthropic. I’m including a couple older models too, like the retired GPT-4.5. This is per million tokens, input / output.\n\n\n GPT-5.5: $5.00 / $30\n GPT-5.4: $2.50 / $15\n GPT-5.4 Mini: $0.75 / $4.50\n GPT-5.4 Nano: $0.20 / $1.25\n GPT-5: $1.25 / $10\n GPT-4.5: $75.00 / $150\n Opus 4.7: $5.00 / $25\n Opus 4.6: $5.00 / $25\n Opus 4: $15.00 / $75\n Sonnet 4.6: $3.00 / $15\n Haiku 4.5: $1.00 / $5\n\n\nWhat can we learn from this? I have no idea. Clearly GPT-4.5 was a gigantic model, and so OpenAI changed course into more efficient models. Opus has generally been more expensive, but now is comparable to GPT-5.5.",
"title": "LLM prices"
}