--- title: Claude Sonnet 5.5 is out, and the early access crowd is... description: Anthropic released Claude Sonnet 5.5, the second model in the Claude 5.5 family, positioned as a clear but faster upgrade over Sonnet 5. Early access user Alex... canonical: https://daily.dev/posts/claude-sonnet-5-5-is-out-and-the-early-access-crowd-is-quietly-impressed-fm8of5mep twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: Claude Sonnet 5.5 is out, and the early access crowd is quietly impressed | daily.dev og:description: Anthropic released Claude Sonnet 5.5, the second model in the Claude 5.5 family, positioned as a clear but faster upgrade over Sonnet 5. Early access user Alex... og:url: https://daily.dev/posts/claude-sonnet-5-5-is-out-and-the-early-access-crowd-is-quietly-impressed-fm8of5mep og:image: https://api.daily.dev/og/posts/FM8OF5mEP.png og:image:alt: Claude Sonnet 5.5 is out, and the early access crowd is quietly impressed og:image:width: 1200 og:image:height: 630 og:locale: en --- > ## Documentation Index > Fetch the complete documentation index at: https://daily.dev/llms.txt > Use this file to discover all available pages before exploring further. # Claude Sonnet 5.5 is out, and the early access crowd is quietly impressed **[Trends](https://daily.dev/sources/trends)** · 3 min read · 3 upvotes · 0 comments ## Summary Anthropic released Claude Sonnet 5.5, the second model in the Claude 5.5 family, positioned as a clear but faster upgrade over Sonnet 5. Early access user Alex Finn described the improvement as subtle rather than immediately obvious, calling it 'bigger news than it appears' despite not noticing a difference at first. Haiku 5.5 is confirmed to follow in the coming weeks, with no specific timeline given, and could matter significantly for high-volume inference use cases given Haiku's speed-and-cost focus. ## Content Anthropic's Claude 5.5 family is getting the kind of reception labs dream about, with one loud asterisk. @theo's verdict: "Sonnet 5.5 is an incredible model. You probably shouldn't use it." He made a video explaining why. Everyone else seems to be busy using it. ## The victory laps The mood is close to relief. @thdxr says Claude is his team's most used model "for the first time since the beginning of the year," crediting Opus 5.5. @trikcode says Opus 5.5 and Sonnet 5.5 "have really restored value to the $200 plan," and that it "feels like 2025, the quality is way higher." @petergyang says Claude Max now feels "basically unlimited on Opus and Sonnet." The jab at OpenAI is explicit. @twostraws sets Anthropic's "unlimited power at incredible speed" against OpenAI "halving how much you get for $200 a month, but here are some temporary credits." He adds that Sol 5.6 Codex "pretty much didn't put a foot wrong", so what happened? ## The party tricks @kloss_xyz asked Sonnet 5.5 on Low effort for a "mind blowing" mascot animation and got 52 seconds of looping pixel art, all in code, from a single prompt. His reaction: "WTF is Anthropic feeding their models bros?" @PawelHuryn threw 105 bugs across two real repos at it and "couldn't believe my eyes." One video shows it reproducing a physics paper (AVBD) and rendering ray-traced caustics. The same video argues each cheaper model now matches its pricier predecessor, so frontier-level AI on laptops is coming. ## The numbers, and the dissent @scaling01 says that even adjusted for task cost, Claude wins: "it's no longer a 5x difference but still 1.3x - 2.9x," with more work done at equal AAI-measured intelligence. Another post flags a step change for Opus 5.5 on CheatBench. Not everyone is dazzled. One reviewer says he can't tell Opus 5.5 apart from earlier models anymore and that it "no longer feels particularly useful." Another praised Dario Amodei, got dunked on for crediting the CEO with model training, and conceded the CEO probably wasn't personally involved, but said the praise stood. ## Meanwhile Vani, a chat and API platform, has already added Sonnet 5.5 alongside Kimi K2.8 Preview and MiMo V2.6. Someone also built a working StarCraft level with it: you play Terran, holding off the Zerg. The pattern: the people paying for the $200 tier are happy, Theo is the contrarian, and the shrugging crowd is the one that has used every model too long to be impressed. ## Questions this post answers ### What is new in Claude Sonnet 5.5 compared to Sonnet 5? Claude Sonnet 5.5 is Anthropic's second release in the Claude 5.5 family, positioned as a clear upgrade over Sonnet 5 with faster performance. Early access tester Alex Finn noted the improvement wasn't immediately obvious but felt more substantial after continued use, describing it as subtle rather than a dramatic leap. Haiku 5.5 is expected within coming weeks. _Developers deciding when to upgrade their Claude integration can track model releases like this on daily.dev._ ### Is Haiku 5.5 available yet from Anthropic? Haiku 5.5 is not yet released; Anthropic has confirmed it will follow Sonnet 5.5 within coming weeks, with no specific date given. Haiku is Anthropic's speed-and-cost optimized tier, so a stronger Haiku 5.5 could significantly benefit teams running high-volume inference workloads. _Teams planning high-volume inference upgrades can follow Haiku model updates on daily.dev._ ## Community take How the wider developer community reacted, aggregated from 1 discussion and 20 comments across x (as of 2026-10-06). **TL;DR:** Discussion centers almost entirely on cost-per-task comparisons between Claude and competing models, with people debating whether the pricing/performance gap is as large as headline numbers suggest. **Sentiment:** 10% positive · 70% mixed · 20% skeptical **The case for** - Several argue cost-per-finished-task is a fairer metric than per-token price, and on that basis the gap to competitors shrinks dramatically. - Some say the throughput/speed advantage matters more day-to-day than raw pricing. - One commenter claims the cost-adjusted price gap closing suggests the broader price war is effectively over in this model's favor. **The pushback** - Others note headline multipliers (like 5x) vary wildly depending on effort level, with real-world ranges closer to 1.3x–2.9x. - Some say their own runs still fail on retries, making cheaper models win once a task needs multiple attempts. - A few point out competing chat products offer near-unlimited usage that offsets the comparison. - One raises concerns about long-context attention dilution causing dropped edge-case constraints in multi-agent setups. **By community** - x (mixed): Replies focus on dissecting cost-per-task vs per-token pricing claims, with genuine disagreement over how large the real-world advantage actually is. **Hottest debate:** Whether the true cost/performance advantage is closer to the headline 5x figure or a much smaller 1.3x–2.9x once task completion and retries are accounted for. **Open questions** - Does the cost-adjusted advantage hold up on long agentic runs with retries and failures? - How does API usage in editors like VS Code/Cursor compare to using the dedicated CLI tool? **Highlights** > @scaling01 Reading the chart, the 2.9x is the max effort row. Low through xhigh come out around 1.3x to 1.7x, so the number you land on depends on which effort level you'd actually run. > — [noelkimx on x · 1 points](https://x.com/noelkimx/status/2107333515163709864) > @scaling01 1.3x to 2.9x after you charge for the task is the number I'd actually test, not the headline 5x. My own runs still blow up on retries, so the cheaper model wins once a step fails twice. > — [chnhn05782803 on x](https://x.com/chnhn05782803/status/2107315702558535993) > @scaling01 per-token pricing is a vanity metric, task cost is the pnl. claude closing a 5x sticker gap to 1.3x means the price war is over and the throughput war is won. still quoting per-token rates is like comparing gas receipts while the other guy owns the refinery > — [EthGoldi on x](https://x.com/EthGoldi/status/2107336472718434534) > @scaling01 Long context windows create an illusion of safety. Once conversations pass 80k tokens, attention dilution causes subagents to drop subtle edge-case constraints. We now enforce sliding semantic contracts between multi-agent handoffs. > — [kiteciyuan48037 on x](https://x.com/kiteciyuan48037/status/2107341269194391702) > @scaling01 Cost charts never show what I actually pay, which is waiting. I will take the pricier model if it finishes before my coffee goes cold. The gap I feel on a Tuesday is minutes, not cents. > — [overshadowed\_x on x](https://x.com/overshadowed_x/status/2107323858005131488) **Source threads** - [x](https://x.com/scaling01/status/2107308460157407267) · 0 points · 20 comments ## Similar posts on daily.dev - [Anthropic Sonnet 5: It closes the gap with Opus 4.8, and is cheap until August](https://daily.dev/posts/anthropic-sonnet-5-it-closes-the-gap-with-opus-4-8-and-is-cheap-until-august-o0toggvty) · The New Stack · 0 upvotes · 0 comments --- Tags: [#llm](https://daily.dev/tags/llm), [#ai-coding](https://daily.dev/tags/ai-coding), [#claude](https://daily.dev/tags/claude), [#anthropic](https://daily.dev/tags/anthropic) [View this post on daily.dev](https://daily.dev/posts/claude-sonnet-5-5-is-out-and-the-early-access-crowd-is-quietly-impressed-fm8of5mep) ```json {"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]} {"@context":"https://schema.org","@type":"TechArticle","headline":"Claude Sonnet 5.5 is out, and the early access crowd is quietly impressed","url":"https://daily.dev/posts/claude-sonnet-5-5-is-out-and-the-early-access-crowd-is-quietly-impressed-fm8of5mep","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/claude-sonnet-5-5-is-out-and-the-early-access-crowd-is-quietly-impressed-fm8of5mep"},"datePublished":"2026-09-29T03:44:34.361Z","dateModified":"2026-10-06T06:15:19.616Z","description":"Anthropic released Claude Sonnet 5.5, the second model in the Claude 5.5 family, positioned as a clear but faster upgrade over Sonnet 5. Early access user Alex...","isAccessibleForFree":true,"articleSection":"Trends","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"Trends","logo":"https://media.daily.dev/image/upload/s--ZfSp3asX--/f_auto,q_auto/v1780996004/logos/trends?_a=BAMAMiWQ0","url":"https://daily.dev/sources/trends"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/claude-sonnet-5-5-is-out-and-the-early-access-crowd-is-quietly-impressed-fm8of5mep","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":3},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"llm,ai-coding,claude,anthropic","timeRequired":"PT3M"} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"Trends","item":"https://daily.dev/sources/trends"},{"@type":"ListItem","position":3,"name":"Claude Sonnet 5.5 is out, and the early access crowd is quietly impressed"}]} {"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/claude-sonnet-5-5-is-out-and-the-early-access-crowd-is-quietly-impressed-fm8of5mep#faq","mainEntity":[{"@type":"Question","name":"What is new in Claude Sonnet 5.5 compared to Sonnet 5?","acceptedAnswer":{"@type":"Answer","text":"Claude Sonnet 5.5 is Anthropic's second release in the Claude 5.5 family, positioned as a clear upgrade over Sonnet 5 with faster performance. Early access tester Alex Finn noted the improvement wasn't immediately obvious but felt more substantial after continued use, describing it as subtle rather than a dramatic leap. Haiku 5.5 is expected within coming weeks. Developers deciding when to upgrade their Claude integration can track model releases like this on daily.dev."}},{"@type":"Question","name":"Is Haiku 5.5 available yet from Anthropic?","acceptedAnswer":{"@type":"Answer","text":"Haiku 5.5 is not yet released; Anthropic has confirmed it will follow Sonnet 5.5 within coming weeks, with no specific date given. Haiku is Anthropic's speed-and-cost optimized tier, so a stronger Haiku 5.5 could significantly benefit teams running high-volume inference workloads. Teams planning high-volume inference upgrades can follow Haiku model updates on daily.dev."}}]} ```