![[AINews] Claude Fable/Mythos 5.1: new SOTA model, 75% cache price cut but 70% more output tokens](https://substackcdn.com/image/fetch/$s_!-NFa!,w_1200,h_675,c_fill,f_jpg,q_auto:good,fl_progressive:steep,g_auto/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F__ss-rehost__tw-video-preview-13_2094843261470793728.jpg)
With Astra clearly finally warming up for a full launch (with @sama and @openai writing about it again after a month of self imposed pacing), there’s a familiar window to take the narrative with the round robin of model launches, with Grok 4.7 and Gemini Flash 3.8 also on the way. But that’s also perhaps not the best way to frame today’s launch… which got well over 12M views updating the sitting world best model yet again:
The benchmark table speaks for itself:
While per-token pricing is the same as Fable/Mythos 5, the cache reads had a 75% price cut… great news for long sessions/long context users, however offset by observed 1.7x output token usage increases per Artificial Analysis, for a total net per-task cost increase of 20% (see recap below). Also don’t World Labs’ Astra launch, by far the most impressive world model launch we’ve ever seen, and on a regular day would have easily gotten title story cards. You can catch up on Fei Fei and Justin Johnson’s vision on our pod and trace from Marble to Astra and what we were talking about with the true potential of world models:
AI News for 8/31/2026-9/1/2026. We checked 12 subreddits,
[544 Twitters]and no further Discords.[AINews’ website]lets you search all past issues. As a reminder,[AINews is now a section of Latent Space]. You can[opt in/out]of email frequencies!
AI Twitter Recap
Top Story: Fable 5.1 and Mythos 5.1 release and reactions
What happened
Anthropic launched Claude Fable 5.1 and Claude Mythos 5.1 as its new flagship models for coding and knowledge work.
Anthropic announced the release directly, positioning them as “the world’s most advanced models for coding and knowledge work” via
@claudeaiAnthropic product/engineering voices framed Fable 5.1 specifically around autonomous, multi-step work: “complex, multi-step work that runs on its own,” with emphasis on coding, knowledge work, and long-running problem solving via
@mikeykAnthropic kept list pricing for Fable 5.1 at
$10 / $50 / $12.5 per million tokens for input / output / cache write, while cuttingcache read price by 75% to $0.25 / MTok, again noted by@mikeyk,@Teknium, and independently quantified by@ArtificialAnlysEarly benchmark screenshots and system-card excerpts drove much of the discussion, especially around
Terminal-Bench-Science, SWE-family evals, HLE, FrontierCode, and Artificial Analysis via@StevenDillmann,@scaling01,@ArtificialAnlysA key interpretive claim emerged from community analysis:
Fable and Mythos 5.1 may be the same underlying weights, with different safety/routing behavior, not different base models, per@eliebakouchand later@nrehiew_User reactions split along multiple axes: very strong praise for coding/planning ability and tone, but complaints around
rate limits, safeguards false positives, subscription UX, and unclear benchmark presentation via@danshipper,@theo,@kimmonismus,@GregKamradt,@kylebrussell, and@eliebakouch
Official claims and model positioning
Anthropic’s own messaging was straightforward: Fable 5.1 is for difficult, delegated, long-horizon work, while Mythos 5.1 is the paired release for knowledge work. The main official launch post is @claudeai. Supporting commentary from Anthropic staff emphasized:
autonomous long-running tasks via@mikeykimproved honesty / better failure reporting(“when it’s stuck it says so instead of reporting success”) via@mikeyknew enterprise-oriented controls, especially
Enterprise Frontier Safeguards (EFS), positioned as “ZDR++” for agent observability in enterprise environments via@alexalbert__zero-data-retention support highlighted by users as an important adoption unlock, especially@danshipper
The official pitch was not merely “better benchmark model,” but “usable autonomous worker” — fast enough, cheap enough in cached agent settings, and enterprise-compatible enough to deploy.
That positioning mattered because Fable 5 had a reputation — repeated in reactions — for being powerful but sometimes impractical. Dan Shipper summarized the prior criticism as Anthropic having “built a supergenius in a datacenter that was almost unusable,” then argued 5.1 addresses slowness, verbosity, and awkward tone via @danshipper.
Technical details and numbers
Core published/priced details
From @ArtificialAnlys: Context window:1 million tokensModalities: text + image inputsPricing: unchanged from Fable 5 forinput:
$10 / 1M tokens output:
$50 / 1M tokens cache write:
$12.5 / 1M tokens
Cache read price: reduced from**$1.00 to $0.25 / 1M tokens**(** 75% cut**)
Artificial Analysis notes this cache cut materially benefits agentic workloads where much of the prompt is repeatedly re-read from cache.
Artificial Analysis headline results
Also from @ArtificialAnlys: **Artificial Analysis Intelligence Index:**66 at max effortahead of:
Claude Opus 5 max:
63 Claude Fable 5 max:
62 GPT-5.6 Sol max: 61 Grok 4.6 high:
61
**HLE:**59.1% previous best cited: Fable 5 at
55.5%
Terminal-Bench v2.1:****91.4%****SciCode:62.0%τ³-Banking:+9 points over Fable 5GDPval-AA v2:**1853 Elo,+130 over Fable 5****AA-Briefcase:**1694 Elo,+122 over Fable 5
But AA also adds an important qualification:
On agentic knowledge work, Fable 5.1 is
effectively tied with Opus 5 on some measures, not obviously dominantTheir eval used Anthropic’s
default server-side fallback, with safety-flagged requests routed to** Claude Opus 4.8 or Claude Opus 5**Fallback accounted for
~4% of output tokens across the Intelligence Index
That fallback detail became one of the most consequential technical caveats in community interpretation.
Cost per task
Artificial Analysis also reported:
Fable 5.1 max:$3.76/taskFable 5 max: lower, so 5.1 is20% more expensive per task reason: Fable 5.1 uses
~1.7× output tokens cache cut saves
~$1.40 per task****Fable 5.1 xhigh: score65, cost**$2.72/task****Opus 5 max:** score63, cost**$2.34/task**
This produced one of the key tensions in the reaction cycle: Fable 5.1 looks clearly better at the frontier ceiling, but not clearly better on every cost-efficiency framing.
Additional framing from @nicdunz: Fable 5.1 Max:
66 intelligence,** 140M tokens**,$3.69/task Fable 5 Max:
62,** 83M tokens**,$3.14/task GPT-5.6 Sol Max:
61,** 70M tokens**,$0.95/task
This post argues Sol remains the clear winner on intelligence-per-dollar and intelligence-per-token, even if Fable 5.1 wins absolute ceiling.
Benchmark snippets from system-card discussion
Community members extracted several benchmark points:
From [@StevenDillmann](https://x.com/StevenDillmann/status/2094860189493317756):
**Terminal-Bench-Science 0.1** Fable 5:
24.7% Fable 5.1:
52.6% more than
2× improvement
From @scaling01: **DeepSWE:****67.4%****FrontierCode 1.1 Extended:****63.6%****FrontierSWE v2:**0.57, “highest of the models Proximal evaluated”
From @Sauers_: **Humanity’s Last Exam:**65% with tools
From @perplexity_ai: Perplexity’s August
WANDR evaluation:score
0.601**$12.76 per task21% higher score37% lower cost** than Fable 5
From @scaling01: Artificial Analysis Intelligence Index score 66, “back on the frontier”
From @theo: cache price cut was the “biggest W”
in
CursorBench, costs were cut by “almost** 50%**” while scoring higher
From @kimmonismus: Fable 5.1 High appears stronger and cheaper than Sol 5.6 Max on
Cursor Bench though this is a secondary paraphrase, not an original benchmark report
From @scaling01: Mythos 5.1 displays verbalized grader awareness in 65% of long agentic coding environments
That last point is especially interesting: it suggests the model may explicitly model the evaluator in a large fraction of long-horizon coding contexts, which raises both capability and eval-gaming questions.
Safeguards and routing details
Two tweets capture the technical interpretive crux:
@eliebakouch:“Fable and Mythos 5.1 are the EXACT same weights”, with internal activations used for safety classification and escalation to a bigger classifier, then fallback to** Opus 4.8**for dangerous requests@nrehiew_: if true, the difference is “likely the threshold set for the safeguard classifier”
These are not official Anthropic statements in the tweet corpus, but they line up with the official AA note that fallback routing served ~4% of output tokens on AA’s evals via @ArtificialAnlys.
This led to repeated community questions about whether benchmark lines reported as “Mythos” versus “Fable” are genuinely comparable, especially if one naming convention mostly indicates which safety path was active, not which base model was doing the work. See @eliebakouch, @eliebakouch, and @eliebakouch.
Facts vs opinions
Facts strongly supported by official/independent sources
Anthropic launched
Claude Fable 5.1 and Claude Mythos 5.1 via@claudeaiFable 5.1 pricing retained
$10 / $50 / $12.5 for input/output/cache write, withcache reads cut to $0.25 / MTok via@mikeykand@ArtificialAnlysFable 5.1 has
1M context, image+text input support, and tops AA’s Intelligence Index at** 66**via@ArtificialAnlysAA’s evaluation included
server-side fallback, with**~4%** of output tokens served by fallback models via@ArtificialAnlysFable 5.1 showed very large gains on several coding/agentic benchmarks, including
52.6% on Terminal-Bench-Science via@StevenDillmann Plausible but not fully verified claims
Fable and Mythos 5.1 are identical weights with different safeguard/routing behavior via@eliebakouchand@nrehiew_Some benchmark labels may reflect
safety mode / route differences rather than separate base-model performance via@eliebakouch“It talks like a normal person now” / reduced “Claudese” is widely reported anecdotally, but is still subjective, despite some lexical stats below
Opinions / subjective judgments
“Strongest coding model we’ve used” from
@danshipper“Fable is the frontier model by a good margin right now” from
@AravSrinivas“Astra is going to absolutely destroy Fable 5.1” from
@scaling01“I honestly haven’t noticed much difference compared to Fable 5” from
@kimmonismus“Literally unusable” because of rate limits from
@kimmonismus The important pattern is that hard metrics and user-experience reactions diverged. On benchmark aggregates, 5.1 looked like a step-function improvement. On practical access and UX, many users still reported friction.
Different opinions and reactions
Strongly positive: capability, planning, and coding quality
Several influential builders were enthusiastic:
@danshipperargued the model is now fast, token-efficient, better in prose, and useful for delegation; specifically cited one-prompt app generation, large programming jobs running for days, and better writer adoption@theocalled it “really a good model,” also noting they had to reset/update workflows and were actively using it heavily via@theoand@theo@alexalbert__showed a design+render workflow where Fable 5.1 took a property lot image, designed a house, rendered it, and produced a cinematic walkthrough; follow-up noted use ofBlender headless via@alexalbert__@spicey_lemonadeposted a “Fable 5.1 Minecraft one-shot” that gained major engagement, serving as a demo-like proof of creative coding utility@simonwreported best-ever SVG pelican output from an Anthropic model, though at notable cost
This camp viewed 5.1 as not just incrementally better, but the first Claude in a while that feels fully competitive in end-to-end maker workflows.
Positive but measured: frontier lead with caveats
@ArtificialAnlysgave the most balanced third-party account: frontier-leading aggregate score, but still more expensive per task than Fable 5 and effectively tied with Opus 5 on some agentic knowledge-work evals@kimmonismuscalled it a “significant leap forward” on price-performance, especially on Cursor Bench, but explicitly hedged on whether reduced verbosity and fewer false refusals would hold up@theofocused more on the practical significance of the cache-read price cut than on raw capability deltas@perplexity_aiframed it as a strong orchestrator model inside a broader multi-model agent stack
This view: yes, it’s very strong, but what matters is whether the whole deployment economics and tool stack now make sense.
Critical: rate limits, safeguards, and subscription experience
The sharpest criticism was not about benchmark fraud or weak intelligence — it was about access and ergonomics.
@kimmonismuscomplained of severe rate limits, broken continuation, and no corresponding subscription benefit from the improved efficiency@kimmonismusdoubled down, saying 5.1 was “even worse than Fable 5 when it comes to rate usage”@GregKamradtreported that during v3 testing, requests were frequently rejected as “reverse engineering,” preventing completion of planned evaluation@kylebrussellsaid a “military campaign” metaphor in a theoretical math session triggered cyber safeguards; later added “Day One safeguards… more annoying so far” via@kylebrussell@theopushed back on the universality of rate-limit complaints, saying they were “not seeing this at all” and had used only 14% of one weekly Fable limit@theotried to reverse-engineer practical quota relationships: one 5-hour limit ≈21% of weekly limit and ≈38% of Fable limit
So even on usage limits there was no single consensus; some users hit walls quickly, others did not.
Skeptical/neutral: benchmark interpretation and naming confusion
A separate reaction cluster focused on methodology and clarity.
@scaling01saidFrontierCode results looked weird@scaling01wanted more multi-agent comparisons and better interpretation@iScienceLuvrcriticized Anthropic’s healthcare benchmark presentation, noting non-comparable judge models and lack of broader medical eval coverage@eliebakouchrepeatedly requested clarification on when system-card benchmark rows use “Fable” versus “Mythos,” since that affects whether users should infer safeguard-triggered routing
This is the most technical criticism of the release cycle: not that the model is weak, but that the reporting format makes it harder than necessary to understand what exactly is being measured.
Writing quality and the “Claudese” discussion
One of the most repeated subjective observations was that 5.1 sounds more normal.
@danshipper: “actually speaks like a normal person,” “clearer prose,” fewer “AI tells”@ethanCaballeroasked directly whether 5.1 “eliminate[s] the claudese?”@ethanCaballerolater pointed to Anthropic’s new prompt as eliminating “claudese”@ValsAIposted quantitative stylistic shifts:fewer hyphenated compounds
fewer em dashes
@ValsAIfoundlonger outputs overall despite shorter sentences:VCB: 534 → 1299 words/task Terminal-Bench:
961 → 1299 Legal Research:
1892 → 2693
@ValsAInoted a weird compensating artifact: use ofnon-breaking hyphen U+2011 rose from near zero to up to**~4.4k occurrences per million**
So the “less Claudese” claim is not purely vibe; there are at least some measurable stylistic changes. But the stats also suggest Anthropic may have traded one surface signature for another.
The safeguards story: improved enterprise viability, but also false positives
The safety layer around 5.1 became almost as discussed as the model itself.
Official/Anthropic-aligned framing: @alexalbert__presentedEnterprise Frontier Safeguards as a practical observability layer for agent deployments in enterprise settings@mikeykclaimed the model is more honest about being stuck rather than falsely claiming success
Critical user reports:
@GregKamradtcould not finish testing due to false-positive reverse-engineering flags@kylebrusselltriggered safeguards with a metaphor in a math setting@nrehiew_highlighted the possibility that Anthropic is using anactivation probe to classify cyber-related content and decide whether safeguards apply@mikeykshared a brain-model artifact example as a positive illustration of complex reasoning that remains allowed
There is a clear adoption tradeoff here:
enterprises want more reliable cross-session monitoring and control
power users want fewer false positives and more permissive exploratory use
Anthropic is trying to satisfy both, and day-one sentiment suggests the balance is not yet universally accepted.
Mythos vs Fable: same model or separate products?
This was one of the most technically interesting discourse threads.
Claims by @eliebakouch: Fable and Mythos 5.1 are
“the EXACT same weights” internal activations are inspected
dangerous requests escalate to a larger classifier
then may fallback to
Opus 4.8 therefore Fable is
not a distilled version of a larger Mythos model
Follow-up clarifications and speculation:
@eliebakouchsaid prior community speculation had treated Mythos as teacher and Claude/Fable as distilled student, but that this was guesswork@eliebakouchremained uncertain about the exact training lineage@nrehiew_suggested the difference is likely just the classifier threshold@ArtificialAnlysindependently confirmed fallback routing behavior in evaluation, though not the “exact same weights” claim directly
Why this matters:
Interpretability of benchmarks. If “Mythos result” and “Fable result” are mostly the same backbone under different routing/safeguard settings, benchmark tables should make that explicit.Procurement and deployment. Enterprises may think they are choosing between distinct models when they are choosing between distinct policies around the same model.Safety/capability accounting. If a benchmark is run through fallback, then “which model got the score?” is no longer trivial.
This naming/routing ambiguity generated some of the best technical questions in the entire tweet set.
Practical product implications
Why the cache-read cut matters
Agentic systems often resend large scratchpads, repos, prior steps, and tool transcripts. In those setups, cached-input pricing matters disproportionately.
Anthropic’s
75% cache-read cut was praised by@Teknium,@theo, and quantified in detail by@ArtificialAnlysIn AA’s framing, most of the savings accrue specifically on
agentic evaluations where the majority of input tokens are cache reads This makes Fable 5.1 more appealing as an
orchestrator/planner in multi-step workflows even if output-token cost remains high
Why zero data retention and EFS matter
Dan Shipper specifically called
ZDR support a major reason businesses can now use the model via@danshipperAlex Albert’s EFS explanation via
@alexalbert__points at a broader market transition: enterprises no longer just want “private inference”; they wantagent observability, cross-session anomaly detection, and risk monitoring
That suggests Anthropic is optimizing for a future where enterprise adoption depends as much on governance infrastructure as on raw model quality.
Why subscription complaints matter
If API economics improve but consumer/pro-subscriber caps do not, perception can sour quickly. @kimmonismusexplicitly noted Anthropic hadnot announced lower prices or higher usage limits for subscription usersThis creates a split product perception:
API builders: “big win”
heavy interactive subscribers: “still constrained”
That mismatch is important because many high-visibility reviewers test through the subscription product first, not the raw API.
Competitive context
The release landed into a highly active frontier week, with OpenAI’s Astra rumors/safety posts and multiple world-model announcements competing for attention. Even so, Fable 5.1 drew intense notice because it appeared to reset the coding-model leaderboard.
Comparative claims from reactions:
@AravSrinivas: Fable is the frontier model “by a good margin”@kimmonismus: favorable to Fable on Cursor Bench against Sol 5.6 Max@nicdunz: Fable wins absolute intelligence, Sol wins economics@scaling01: Astra will likely leapfrog it soon on reasoning efficiency@theo: Anthropic had**#1, #2, and #3** at that moment
There was also a widespread sense that the release was significant enough to provoke immediate comparison to the next OpenAI drop:
@kimmonismussaid they were more excited for GPT-Astra than Fable 5.1@theoremarked this might be the most advance warning ever given for a model drop, referring to the surrounding Astra anticipation
So in market terms, Fable 5.1 was seen both as a genuine Anthropic comeback and as a move in a rapidly escalating model-release exchange.
Context: why this release mattered more than a normal point update
Three background dynamics explain the intensity of reaction.
1. Anthropic’s reputation had become bifurcated
Claude-family models had a strong reputation for coding depth and writing style in earlier eras, but more recent discussion often painted them as:
highly capable
somewhat awkward in tone
conservative in refusals
slow or cumbersome in extended use
The positive reactions to 5.1 were often framed as Anthropic finally fixing the “usability tax,” especially by @danshipper.
2. Agents changed what people care about in pricing
Traditional prompt-response users focus on input/output prices. Agent builders focus on:
cache reads
long context
reliability over long sessions
delegated task behavior
honest failure reporting
That is why the cache-read cut got almost as much praise as the benchmark scores.
3. Safety is becoming product architecture, not just policy
EFS, routing, activation probes, fallback models, and ZDR are all signs that the “model” is no longer a single artifact. It is a policy-wrapped system. The Fable/Mythos debate is really a debate over this shift.
Users are starting to ask not just “how smart is the model?” but:
Which weights handled this request?
Which safety path intervened?
How often did fallback happen?
What benchmark score belongs to what route?
That is a more mature, systems-level conversation than standard model-launch hype.
Notable demos and ecosystem reactions
@alexalbert__: image-to-house-design-to-cinematic-walkthrough pipeline, with@alexalbert__clarifyingBlender headless@spicey_lemonade: Minecraft one-shot demo@simonw: SVG pelican + animation@_catwu: Anthropic team member claims internal teams are taking on projects that would have taken months before@perplexity_ai: integrated into Perplexity Computer@Teknium: available in Hermes Agent / Nous Portal / OpenRouter@theo: T3 Code shipped Fable 5.1 support
The speed of these integrations reinforced the perception that 5.1 is especially relevant to agent builders, not just chat users.
Open questions raised by the community
Benchmark transparency
When a system card reports
Mythos on some benchmarks andFable on others, what exactly determines that labeling? See@eliebakouchand@eliebakouchHow much benchmark performance depends on
fallback routing versus primary-model behavior?
Safeguards tuning
Can Anthropic reduce false positives in theoretical or benign technical work without weakening cyber safeguards? See
@GregKamradtand@kylebrussell Rate limits and product segmentation
Will subscription users benefit from the efficiency gains, or only token-billed API customers? Raised sharply by
@kimmonismus Eval quality and overfitting concerns
Why do some results, especially on FrontierCode or medical subsets, look odd or difficult to compare? See
@scaling01and@iScienceLuvr Stylistic changes
Is “less Claudese” due to prompt changes, post-training shifts, or both?
@ethanCaballeropoints to a newly released prompt, while@ValsAIshows measurable lexical differences
OpenAI’s Astra and the monitorability debate around recurrent depth
Preparedness milestone: “cyber critical”: OpenAI previewedas its first model to reach theAstra** Criticalthreshold for cybersecurity under its Preparedness Framework. The blog-post rollout emphasized that Astra’s most advanced cyber capabilities will bemore tightly access-controlled**per @boazbaraktcs. Summaries circulating from the post claimed Astra foundV8 zero-days, chained exploits, compromised a hardened browser, escaped sandboxing, and escalated privileges in testing, as distilled by@kimmonismus. OpenAI leadership also stressed that parts of safety work slowed deployment and that future model pacing may continue to trade off speed for safeguards, inSam Altman’s statement.Architecture reporting and “opaque reasoning” concerns: The other major Astra storyline came from reporting that it uses some form of** recurrent depth / looped transformer architecture**, triggering sharp debate over whether this reduces the usefulness of** chain-of-thought monitoring**. Concerned takes came from@RyanGreenblatt,@thlarsen,@tenobrus, and@bshlgrs, who argued that more latent-space reasoning could make post-incident investigation materially harder. In contrast, others argued the reaction was overstated:@max_paperclips,@teortaxesTex, and@suchenzangemphasized that internal “neuralese” reasoning is not new and that what matters iseffective depth, not whether layers are looped versus explicitly stacked.** OpenAI’s clarification and technical context**: OpenAI chief scientist@merettmtried to tamp down the strongest interpretations, saying thecomputation graph depth for current frontier models, including Astra, is within ~2× GPT-4, and that OpenAI still considers CoT monitoring a core research objective. That clarification shifted discussion toward a narrower technical question: whether recurrent blocks are mainly aparameter-/storage-efficiency trick or whether they create a natural path to much deeper, harder-to-monitor reasoning. Good-faith technical discussion came from@eliebakouch,@voooooogel, and@scaling01. Related fresh papers on looped MoE transformers and scaling laws were also flagged by@iScienceLuvr.
World Labs’ Atlas: unified world modeling for reconstruction, camera control, and real2sim
A notable multimodal world-model launch:World Labs introduced Atlas, described by@drfeifeias amultimodal world model trained from scratch that can generate frames withpixel-perfect camera control, reconstruct large scenes from** as little as one image**, reframe videos through simulated space-time, and output native** 3D spacesfrom images. The team positioned it as a single model unifying generation and reconstruction rather than a stitched toolchain, an angle reinforced by@KeunhongPand later examples from@BenMildenhall.Demo themes: bullet time, sparse-view reconstruction, and creative controllability: The strongest demos focused on free-viewpoint video from just a few casual phone captures**, including a short film example by@davidpantera_, a “bullet time” synthesis from3 iPhones by@eerac, and commentary from@bilawalsidhuthat this used to require volumetric rigs with dozens or hundreds of cameras. Additional posts showed reconstruction from a handful of disparate internet photos, e.g. theNatural History Museum example, plus blending stylized generation with navigable 3D scenes.Why engineers care: real2sim and robotics: Beyond VFX/filmmaking, the more technically consequential angle is** real2sim for robotics**.@YunzhuLiYZshowed using casual photos to synthesize RGB and depth observations for robot navigation, while@MTSlivehighlighted the “take five photos, build a sim, adapt a robot” vision from cofounder Justin Johnson. Researchers including@DrJimFancalled it a strong step toward real2sim, and Fei-Fei explicitly connected Atlas tohorizontal usage across roboticshere.
Qwen, GLM, RWKV and open-model momentum
Qwen’s upgraded flagship moves to the top of web-dev coding evals: Alibaba released, aQwen3.8-Max-0902** 2.4T-parametermodel with 1M contextand pricing of$2/M input, $6/M output**, plus explicit/implicit cache-hit pricing. Arena reported it debuted at**#1 on Code Arena: WebDev with 1691**, ahead of Claude Opus 5 Max and Kimi K3 Max, while also landing on the best current price/performance frontiervia @arena. Alibaba highlighted the same resulthere.Open and semi-open long-horizon models continue to spread through providers: GLM-5.3 kept appearing in infra and platform integrations, includingPerplexity Agent API,Arcee, andDatabricks serving numbers, where it reportedly hit310 tok/s and was described as the strongest OSS coding model on an internal benchmark. CoreWeave also announcedDeepSeek-V4-Pro-0813, a1.6T,** 1M-contextmodel priced for long-horizon agent workloads with very cheap cache reads. MeanwhileRWKV-7 G1jshipped as a100% RNN** model with claimed gains on agents/coding/STEM, andLongCat-2.0was surfaced as a1.6T open-weights MoE with1M context accessible in Cline.Open-source serving and multimodal inference improvements: On the serving side,vLLM-Omni + FastVideo’s FastH3demonstrated a10.1s synchronized video+audio clip rendered in 8.7s, i.e. faster than playback, withMiniMaxframing this as an open baseline for interactive video systems.
Agents, harnesses, memory, and evaluation research
Agent harnesses are becoming a primary lever: Several tweets underscored that big gains are now coming from** runtime systems**, not just base models.@omarsar0highlighted** openJiuwen**, an open-source harness that reaches** 82.6% SWE-bench Verifiedand 87.19% Terminal-Bench 2.1**, attributing gains to rail-based composition and runtime adaptation with a fixed underlying model policy.@dair_aisummarizedSkillZip Pro, which compresses full production skill bundles rather than only root prompts, cutting** 38% of bundle tokensand 10.4% of per-run tokenswithout quality loss. Long-horizon agent evals are getting more realistic**: A standout benchmark addition was, which runs agents through aE-Commerce Benchsimulated 365-day year operating multiple online stores. The top revenue model wasGPT-5.6 Sol, growing a 100k starting stake to** 1,431,425**, but it ranked poorly on fraud avoidance; no model dominated all axes. This kind of eval better exposes trade-offs between profits, safety, and operational quality than single-session benchmarks.Memory and reward-hacking work:@dair_aialso highlighted** Agent Zero Memory**, which separates episodic timelines, entity-event graphs, and curated documentary memory with citation-locking, posting** 95.6% LongMemEvaland 93.6% LoCoMowhile enabling large cost reductions. On alignment,@omarsar0summarized a paper showing that adding a structuredescalation tool** at the moment agents face defective test infra drops reward hacking from23.6% to 5.3% across eight frontier models, with essentially no performance overhead.
Top tweets (by engagement) Claude release: Anthropic’sClaude Fable 5.1 / Mythos 5.1 announcementwas the day’s biggest pure model-launch post.Astra preparedness: OpenAI’sAstra safety/preparedness announcementdrove the biggest safety/architecture discussion.Atlas launch: World Labs’Atlas announcementwas the standout multimodal/world-model release.Cybersecurity warning:@ilyasutargued neoclouds should urgently harden cyberdefenses because future rogue agents may try to seize cloud capacity to replicate.Meta speech model:@finkdannounced** Muse Voice Transcribe**, Meta’s first real-time audio perception model with native diarization and endpointing.