Fable 5 costs$10 per million input tokens and $50 per million output tokens.

Opus 5 costs$5 and $25.

Give each model one million input tokens and one million output tokens, and the bill is simple:

Opus 5: $30
Fable 5: $60

Both can take a one-million-token context. Both are built for serious coding, research and agent work. Yet the cheaper model leads several of Anthropic’s published evaluations for computer use, business automation, search and general agentic work. At maximum effort, Anthropic says Opus 5 comes within 0.5% of Fable 5’s best CursorBench result while costing half as much per completed task.

That makes this comparison much more useful than the usual “which model is smarter?” argument.

The real decision is:

Which model should think about the job, which one should perform it, and how much effort should each part receive?

That is where the savings are.

The result I did not expect

I tested both across the work I actually care about: planning, coding, research, long prompts, tool use and repeated agent loops.

Fable 5 did not feel like a model that should be used for everything. It felt like a model that should be brought in when theplan itselfwas the dangerous part.

Opus 5 was easier to keep moving. It was strong enough for the large middle of real work: inspecting a codebase, implementing features, using tools, browsing, debugging, creating files and fixing its own output.

Fable was more useful when I wanted the model to sit with a difficult problem before touching anything: architecture, a large migration, an unclear investigation or a task expected to remain autonomous for hours or days.

My working rule became:

Fable plans. Opus builds. Fable reviews only when the result is expensive to get wrong.

Anthropic describes Opus 5 as the everyday model and Fable 5 as the model for ambitious, long-running projects. Fable can work across stages for days, delegate to subagents and check its work with limited supervision.

The difference in one table

Opus wins several large published comparisons, including Frontier-Bench, OSWorld, AutomationBench, BrowseComp and GDPval-AA. Fable retains smaller leads on evaluations such as DeepSWE, SWE-bench Pro and some legal or long-horizon work. The narrow differences should not decide a production system on their own.
Your own acceptance test matters more than one leaderboard row.

If you want the individual breakdowns before going deeper, I’ve already published separate guides on both models

Inside the full guide: the exact Opus 5 vs Fable 5 setup, effort routing, Claude Code commands, agent loops, Fable planning, Opus execution, Graphiti memory, reusable skills, task budgets, prompt templates, and the hidden review and hallucination risks.

",{"_244":190,"_245":273,"_246":274},"dd7bca0b-2e5e-4a0a-a45d-eac9b9ecddc0",{"_19":273,"_31":275,"_192":276,"_33":277,"_215":278,"_251":278,"_21":222,"_252":279,"_254":280,"_208":40,"_68":-5,"_70":40,"_170":281,"_244":190,"_270":294},"Opus 5 >> Fable 5","The article discusses the potential of tldraw's offline app for creating visual agents and interactive experiences, moving beyond text-based AI interactions. It covers recent advancements in large language models, including Anthropic's Claude Opus 5 and ChatGPT's new voice capabilities, highlighting both their strengths and limitations. Additionally, it touches upon various AI tools and platforms, such as FLUX 3, BackSearch, and open-source projects like Ability AI's Trinity, emphasizing the growing trend of agent-centric development and visual interfaces.","Codex plus Voice is the new openclaw","https://substackcdn.com/image/fetch/$s_!bJlF!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fpbs.substack.com%2Fmedia%2FHORSNyKXMAA28Yh.jpg","2026-07-28T13:10:18","https://www.bensbites.com/p/opus-5-fable-5",{"_19":282,"_173":283,"_175":115,"_68":284,"_201":291,"_202":292,"_204":293},"e377c198b51ace7c37d46fc02006462b7afa6b0b51ba54169da1aa417516450a","2190f40ab7a747b95171148c4d82965a0d38922c469064c1b11bba36e0ad95153f66805ec2e547724414d5597cc9526670564913cd74deff5bed63195c3b7cb2",[285,286,287,289,290],[208,275],[35,278],[263,288,40,265],"30023:0b16efb3db6165f2311a31e04c8f874f8c569a4cc6cc219a3295cdfa9110b8b8:dd7bca0b-2e5e-4a0a-a45d-eac9b9ecddc0",[198,40],[200,291,40],"0b16efb3db6165f2311a31e04c8f874f8c569a4cc6cc219a3295cdfa9110b8b8","Opus 5 >> Fable 5\nCodex plus Voice is the new openclaw\n\n\nnostr:naddr1qvzqqqr4gupzqzcka7eakct97gc35v0qfj8cwnuv26dye3kvyxdr99wdl2g3pw9cqqjxgephvf3kzvrz95ex2dt9956xzvrp94sngdty94jkzceevguk2cmyv33nqw3mj6j ",1785244218,"

Hey folks,

I’m back from holiday and straight back tomessing aroundbuilding. I’ve been kinda obsessed and enamoured with the demos from tldraw. It’s a free canvas/whiteboard app but their offline app (recentlyupdated) is cool because agents can do all sorts with it; create visualisations to explain stuff, make games, interactive blog posts and a ton of other stuff.

Like turning our logo into a little animating mascot. He’s called ‘bites’ btw.

or little widgets like a timezone checker

@tldrawoffline \",\"username\":\"bentossell\",\"name\":\"Ben Tossell\",\"profile_image_url\":\"https://pbs.substack.com/profile_images/1878086921726943233/vOx1kjeP_normal.jpg\",\"date\":\"2026-07-27T11:40:18.000Z\",\"photos\":[{\"img_url\":\"https://substackcdn.com/image/fetch/$s_!7Ccf!,w_1028,c_limit,f_auto,q_auto:best,fl_progressive:steep/l_play_button_usfui2,w_88,e_colorize:0/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F__ss-rehost__tw-video-preview-13_2081706218259636225.jpg\",\"link_url\":\"https://t.co/MMiGh3VMAS\"}],\"quoted_tweet\":{},\"reply_count\":2,\"retweet_count\":1,\"like_count\":9,\"impression_count\":11165,\"expanded_url\":null,\"video_url\":\"https://video.twimg.com/amplify_video/2081706218259636225/vid/avc1/428x424/ROZL3DMQl4PuoLl8.mp4\",\"video_preview_media_key\":\"13_2081706218259636225\",\"belowTheFold\":false}\" data-component-name=\"Twitter2ToDOM\">

or a daily weather timeline

@tldraw\",\"username\":\"bentossell\",\"name\":\"Ben Tossell\",\"profile_image_url\":\"https://pbs.substack.com/profile_images/1878086921726943233/vOx1kjeP_normal.jpg\",\"date\":\"2026-07-27T12:09:52.000Z\",\"photos\":[{\"img_url\":\"https://substackcdn.com/image/fetch/$s_!A46f!,w_1028,c_limit,f_auto,q_auto:best,fl_progressive:steep/l_play_button_usfui2,w_88,e_colorize:0/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F__ss-rehost__tw-video-preview-13_2081713656362876928.jpg\",\"link_url\":\"https://t.co/Wi3w9F5mBp\"}],\"quoted_tweet\":{\"full_text\":\"timezone checker in @tldraw offline\",\"username\":\"bentossell\",\"name\":\"Ben Tossell\",\"profile_image_url\":\"https://pbs.substack.com/profile_images/1878086921726943233/vOx1kjeP_normal.jpg\"},\"reply_count\":1,\"retweet_count\":2,\"like_count\":37,\"impression_count\":7078,\"expanded_url\":null,\"video_url\":\"https://video.twimg.com/amplify_video/2081713656362876928/vid/avc1/712x480/rM3MO8lKfitLiFPy.mp4\",\"video_preview_media_key\":\"13_2081713656362876928\",\"belowTheFold\":false}\" data-component-name=\"Twitter2ToDOM\">

I’m not good using agents when everything is just text in a file. It’s boring, I don’t read it and I sure as shit don’t do enough (anything?) with it. So if it’s visual and all on one canvas, maybe I will?

My idea here is to create a little ‘one-stop’ canvas for some widgets I use regularly as well as showing me my todos/emails to sort and things like that. I’m then planning on making my personal site an interactive canvas too, then some more explainer-like stuff too.

And maybe I’ll get the mascot hooked up to an agent so i can drag him to my todos and ask him to do stuff…

I’ll share more soon

Ben’s Bites is brought to you byAbility AI

This startup runs on17 AI agents- and open-sourced the platform.Marketing, client health, competitive intel, finance - real ops, not demos, with every agent action git-tracked and a human approving anything external. Agents you own, not rent. Apache 2.0:github.com/Abilityai/trinity

Headlines

Claude Opus 5 is out- Anthropic says it gets close to Fable 5 at half the price. Claire hated using Opus 5, but when asked to rank model outputs in a blind test,she ended up rating it first.

Every’s reviewlists some behaviours that might explain this: it argues, stops early and fights the skills people built for older models.

And maybe we are overprompting it.

Anthropic says it removed more than 80% ofClaude Code’s system promptfor Opus 5 and Fable 5 with no measurable coding-eval loss.

ChatGPT Voice can now run the ChatGPT desktop app. You can use it in ChatGPT Work and Codex both; just start a “Voice” session and speak to it. It can create new sessions for individual tasks and then report the results to you in the main thread. It feels a lot like OpenClaw with voice.

I didn’t have a great time using ChatGPT Voice for “serious” work. The delegation of tasks to new sessions via ChatGPT Voice doesn’t feel as smart as doing it in a chat with 5.6 Sol in Codex. Maybe it’s the lack of control on what models/thinking level is used for each new session.
I imagine Voice to work great for productivity type tasks, checking emails, creating draft replies, pull some info from a dashboard and create a chart etc.
For serious work, I still prefer using a speech to text app (currently using one I built —optionafk.com) and sending that to the agent of my choice with full control over model/thinking effort.
— Keshav

Also, I wouldn’t blame you if you missed it, butClaude also has a voice mode. Till now, it could only use Haiku, but now it can use Sonnet and Opus as well as call multiple tools like Gmail, Calendar and Slack mid-conversation. I don’t remember when I last used Claude’s chat option. It’s been Claude Code for a long time now.

Kimi K3’sweightsandtechnical reportare out. It’s already available inDroid at 50% off until 10 August.

Following rumours that the US might ban companies from using Chinese open-weight models, NVIDIA’s CEO Jensen Huangmade his first-ever post on X- a long statement backing open-source AI. Over the weekend,everyone signed that statement(OpenAI, Google, Microsoft and 150+ more) - except Anthropic.

Of course, that led to some backlash, and Anthropic did what it does:published an essayto save face, saying it never pushed for an open-weights ban. The reaction to it is people calling Anthropic out on their hypocrisy — they are okay with open models as long as they arecute little 8b toys, but not if they are “big, dangerous models” (to Anthropic’s business).

@AnthropicAI\\\"Only we are capable of determining what ideas are dangerous. Peasants.\\\"\\n~ \",\"username\":\"NachoQuixotic\",\"name\":\"Nacho Business\",\"profile_image_url\":\"https://pbs.substack.com/profile_images/1681772848317100036/B3NcCduH_normal.jpg\",\"date\":\"2026-07-27T23:08:51.000Z\",\"photos\":[{\"img_url\":\"https://pbs.substack.com/media/HORSNyKXMAA28Yh.jpg\",\"link_url\":\"https://t.co/jSxJkE1eVp\"}],\"quoted_tweet\":{},\"reply_count\":4,\"retweet_count\":32,\"like_count\":1283,\"impression_count\":34240,\"expanded_url\":null,\"video_url\":null,\"video_preview_media_key\":null,\"belowTheFold\":true}\" data-component-name=\"Twitter2ToDOM\">

Subscribe now

Quick links

FLUX 3- one model for image, video, audio and action prediction; video is in early access.
Health in ChatGPT- US users can optionally connect Apple Health and supported medical records, then ask about changes in context.
BackSearch- search the web as it was on a particular date in the past. Right now, it only works for news in 2026.
here.now Workspaces- private team home for the sites, docs and dashboards your agents publish.
Why software factories fail -part 1,part 2,part 3.
open-git- a code host built for agents, with an autopilot that fixes your PR when CI goes red.
OpenWorker- open-source Mac agent that finishes everyday work and asks before consequential actions.
Pilot Protocol- a network of over 250,000 agents that you can use to gain distribution for your app. (I’m an investor)
Notion as code- define your whole workspace in TypeScript and deploy it through the API. (beta)
The Stack v3- open code dataset with about 5T filtered tokens across 770 languages.
box- full VM sandboxes for your agents at $0.00001 a second, with snapshots, SSH and a desktop.
Mintlify- turns resolved support tickets into proposed documentation fixes.
Cognition acquired Interaction- the team behind Poke joins the makers of Devin.

Skills section

Interface Craft- four new lessons on advanced animation, plus an updated DialKit skill for timelines. (tweet)
/prototype- asks your agent for several UI versions and saves the chosen one in a shareable URL. (example run)
Transitions.dev- 33+ polished UI transitions as copyable CSS or React, or an installable agent skill. (tweet)
sharechat- share a Claude Code or Codex session as an unlisted link; no signup, no login to read it.

Afters

@PjoyAI@lakshyaagI thought the way to announce such a thing was not to change your bio but to post the 10 paragraph essay that i just shared with the team?\",\"username\":\"karpathy\",\"name\":\"Andrej Karpathy\",\"profile_image_url\":\"https://pbs.substack.com/profile_images/1296667294148382721/9Pr6XrPB_normal.jpg\",\"date\":\"2026-07-26T01:51:27.000Z\",\"photos\":[],\"quoted_tweet\":{},\"reply_count\":119,\"retweet_count\":82,\"like_count\":3007,\"impression_count\":263923,\"expanded_url\":null,\"video_url\":null,\"video_preview_media_key\":null,\"belowTheFold\":true}\" data-component-name=\"Twitter2ToDOM\">

Share Ben's Bites

Find me onX,Linkedin, orYouTube
Readabout meand Ben’s Bites
thumbnail via@keshavatearth

* sponsors who make this newsletter possible :)
Wanna partner with us for the next quarter?
Email us atshanice@bensbites.comork@bensbites.com

Ben's Bites is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.

",[],"selectedArticleImage","rootArticleImageUrl","consensusHtml","tweets",[],"userSchema","relatedArticles",[247,274],"actionData","errors"]

Anthropic's Opus 5 and Fable 5 Models Compared in User Tests

Story coverage