https://tokensboutique.com/blog Notes and updates from Tokens Boutique. en https://tokensboutique.com/blog/100-models-8-new-vendors https://tokensboutique.com/blog/100-models-8-new-vendors Thu, 20 Aug 2026 00:00:00 GMT Steven We just crossed **100 models** on the table, across 19 vendors. Here's everything new. feat: 8 new vendors joined: **Cohere**, **AI21 Labs**, **NVIDIA**, **Amazon**, **MiniMax**, **ByteDance**, **Baidu**, and **Tencent**. feat: Cohere brings **Command A** (``command-a``), AI21 brings **Jamba Mini**… https://tokensboutique.com/blog/new-domain-tokensboutique-com https://tokensboutique.com/blog/new-domain-tokensboutique-com Thu, 20 Aug 2026 00:00:00 GMT Steven Tokens Boutique now lives at [tokensboutique.com](https://tokensboutique.com). The old address, tokens.boutique, redirects there automatically. Bookmarks, old blog links, anything pointing at the .boutique domain will still get you here. Same site, same data, same calculator. Just a domain that's ea… https://tokensboutique.com/blog/were-on-bluesky-now https://tokensboutique.com/blog/were-on-bluesky-now Fri, 14 Aug 2026 00:00:00 GMT Steven If you followed us on X for release notes and pricing updates, here's where that's going now. feat: Tokens Boutique is now on **Bluesky** at [@tokensboutique.bsky.social](https://bsky.app/profile/tokensboutique.bsky.social). New models, price changes, and site updates get posted there first. fix: Th… https://tokensboutique.com/blog/glm-5-3-released https://tokensboutique.com/blog/glm-5-3-released Fri, 14 Aug 2026 00:00:00 GMT Steven Z.ai released **GLM-5.3** today, and for once we can't just drop a rate into the table. feat: Added ``glm-5.3`` to the calculator and analyzer. 1M context, 128K max output, tagged frontier. feat: There's no metered per-token API price for it. Access goes through the GLM Coding Plan subscription only… https://tokensboutique.com/blog/release-of-gemini-3-7-flash-grok-4-6-and-price-cuts https://tokensboutique.com/blog/release-of-gemini-3-7-flash-grok-4-6-and-price-cuts Thu, 13 Aug 2026 00:00:00 GMT Steven Price cuts, new models, and a couple of overdue fixes. feat: Google cut ``gemini-3.6-flash`` in half, $1.50 to $0.75 input and $7.50 to $3.75 output, cached input and storage following the same ratio. It's introductory pricing through 2026-12-31, stepping back up to the original rate on 2027-01-01.… https://tokensboutique.com/blog/alibaba-joins-with-the-qwen-lineup https://tokensboutique.com/blog/alibaba-joins-with-the-qwen-lineup Mon, 03 Aug 2026 00:00:00 GMT Steven New vendor day. Alibaba now has a spot on the site, with four models from their Qwen family. feat: Added **Qwen3.8-Max** (``qwen3.8-max``), released today. $2.00 input, $0.25 cached input, $6.00 output per million tokens. 1M context, 128K max output. feat: Added **Qwen3.7-Max** (``qwen3.7-max``), $1… https://tokensboutique.com/blog/meta-joins-with-muse-spark-1-1 https://tokensboutique.com/blog/meta-joins-with-muse-spark-1-1 Sun, 02 Aug 2026 00:00:00 GMT Steven New vendor day. Meta now has a spot on the site, starting with their first model on the [Meta Model API](https://dev.meta.ai/docs/getting-started/overview). feat: Added **Meta** as a vendor, with their debut model **Muse Spark 1.1** (``muse-spark-1.1``). feat: Priced at $1.25 input, $0.15 cached inp… https://tokensboutique.com/blog/insights-public-preview https://tokensboutique.com/blog/insights-public-preview Sat, 01 Aug 2026 00:00:00 GMT Steven Every price on this site has always been public, on purpose. But picking a vendor to actually build on takes more than a spreadsheet of per-token rates, and that's the gap Insights was built to close. feat: [**Insights**](https://tokensboutique.com/insights) is live now, in public preview. Think ran… https://tokensboutique.com/blog/anthropic-and-spacexai-lineup-updates https://tokensboutique.com/blog/anthropic-and-spacexai-lineup-updates Sat, 01 Aug 2026 00:00:00 GMT Steven A batch of legacy Anthropic and SpaceXAI models joins the table today, alongside one brand new model. No price changes in this one. feat: Added **Claude Mythos 5**, invitation-only through Anthropic's Project Glasswing program. It shares Claude Fable 5's specs and pricing exactly, $10 input, $50 out… https://tokensboutique.com/blog/gpt-5-6-terra-and-luna-price-cut https://tokensboutique.com/blog/gpt-5-6-terra-and-luna-price-cut Sat, 01 Aug 2026 00:00:00 GMT Steven OpenAI cut prices on two of the three GPT-5.6 models on July 30, three weeks after launch. feat: ``gpt-5.6-terra`` goes from $2.50 to $2.00 per million input tokens and $15.00 to $12.00 output, a 20% cut. Cached input follows the same ratio, now $0.20. feat: ``gpt-5.6-luna`` gets the steeper cut, $1… https://tokensboutique.com/blog/cache-read-write-tokens-in-the-calculator https://tokensboutique.com/blog/cache-read-write-tokens-in-the-calculator Sun, 26 Jul 2026 00:00:00 GMT Steven Prompt caching pricing is messy enough that guessing at it was never going to cut it for long. This update lets you paste in the exact numbers instead. feat: The calculator has a **Cache** section now, tucked behind an arrow next to Vendor and Search. Open it and you get two fields, **Cache read tok… https://tokensboutique.com/blog/retired-models-get-their-own-marker https://tokensboutique.com/blog/retired-models-get-their-own-marker Sat, 25 Jul 2026 00:00:00 GMT Steven Deprecated and retired aren't the same thing. A deprecated model still answers API calls, with the vendor telling you to move off it. A retired one returns an error no matter what you send. Until today we only tracked the first one. feat: Retired models now show a plain † next to their name instead… https://tokensboutique.com/blog/claude-opus-5-is-now-live https://tokensboutique.com/blog/claude-opus-5-is-now-live Fri, 24 Jul 2026 00:00:00 GMT Steven Anthropic released **Claude Opus 5** today, and it's already on the site. feat: Added ``claude-opus-5`` to the calculator and analyzer. Pricing is unchanged from Opus 4.8, $5 per million input tokens and $25 per million output, but Anthropic says it closes most of the gap to Fable 5 at half the cost… https://tokensboutique.com/blog/google-gemini-3-6-flash-and-3-5-flash-light-just-landed https://tokensboutique.com/blog/google-gemini-3-6-flash-and-3-5-flash-light-just-landed Wed, 22 Jul 2026 00:00:00 GMT Steven Google shipped two new Gemini models, and they're already live on Tokens Boutique. feat: Added ``gemini-3.6-flash``, their new workhorse model built on Gemini 3.5 Flash with better token efficiency for coding and multimodal tasks. feat: Added ``gemini-3.5-flash-lite``, their cheapest current-generat… https://tokensboutique.com/blog/deeplink-support-now-available https://tokensboutique.com/blog/deeplink-support-now-available Wed, 22 Jul 2026 00:00:00 GMT Steven Found the perfect combination of models, tokens, and filters? Now you can send it to someone else exactly as you left it. feat: Every filter, token amount, and sort order you set now lives in the page's link. Click a model's name and that exact link, tokens and all, copies straight to your clipboard… https://tokensboutique.com/blog/vendor-logos-everywhere https://tokensboutique.com/blog/vendor-logos-everywhere Mon, 20 Jul 2026 00:00:00 GMT Steven Today we are introducing a visual and important update. Every vendor now has their logo assigned next to their name. feat: Pick a vendor and its logo shows in the selector field. , already breaking records! Time for us to do our part and have them listed on our site. feat: We have added the popular new ``kimi-k3`` model to the calculator and analyzer! feat: Also added ``kimi-… https://tokensboutique.com/blog/making-ai-model-pricing-more-transparent https://tokensboutique.com/blog/making-ai-model-pricing-more-transparent Sat, 18 Jul 2026 00:00:00 GMT Steven The more vendors we add, the more it's become clear that a lot of real-world pricing doesn't fit cleanly into "$X per million input tokens, $Y per million output tokens." This release ``v26.1`` is about showing you the facts when calculating estimated costs. feat: On the site and in the analyzer, an… https://tokensboutique.com/blog/cache-hit-pricing-and-a-prompt-analyzer https://tokensboutique.com/blog/cache-hit-pricing-and-a-prompt-analyzer Thu, 16 Jul 2026 00:00:00 GMT Steven Here we are with version ``v26.0`` and a huge update! A lot of new features and fixes. feat: New **Prompt Analyzer**. Paste any prompt and it shows you a token-by-token breakdown, an estimated token count, and the cheapest models for that exact prompt across every vendor. Everything runs locally in… https://tokensboutique.com/blog/tagging-and-hiding-deprecated-models https://tokensboutique.com/blog/tagging-and-hiding-deprecated-models Wed, 15 Jul 2026 00:00:00 GMT Steven feat: Models that vendors have deprecated or superseded are now tagged with a small red dot next to their name. feat: Added a **No deprecated** toggle next to Frontier only, so you can hide deprecated models entirely. fix: Small accessibility improvements to the toggle switches. https://tokensboutique.com/blog/frontier-toggle-and-more https://tokensboutique.com/blog/frontier-toggle-and-more Tue, 14 Jul 2026 00:00:00 GMT Steven A smaller round of changes, mostly about making the table easier to filter. feat: Added a **Frontier only** toggle. Flip it on and the table narrows down to just the current flagship model from each vendor. Useful if you just want to compare the best of the best. feat: The toggle now shows a live co… https://tokensboutique.com/blog/recent-fixes-and-features https://tokensboutique.com/blog/recent-fixes-and-features Tue, 14 Jul 2026 00:00:00 GMT Steven A batch of changes went in today. Here's what's new. feat: The output token field in the calculator is now capped at 384k and the model table can be sorted by max output. feat: Sorting also defaults to most capable first, then cheapest. feat: All prices shown are cache-miss prices. We plan to add ca… https://tokensboutique.com/blog/welcome https://tokensboutique.com/blog/welcome Mon, 13 Jul 2026 00:00:00 GMT Steven Tokens Boutique is a small, fair, and transparent AI token price comparison site.