Meta says Muse Spark 1.3 has frontier efficiency — however its finest outcomes come from a mannequin builders can’t broadly use but

Meta’s latest AI mannequin Muse Spark 1.3, unveiled yesterday, is quicker and extra performant on third-party benchmarks than its predecessor — with a caveat.

“Muse Spark 1.3 is rolling out at present with frontier efficiency nearly too low cost to meter,” Meta co-founder and CEO Mark Zuckerberg wrote on X, calling it Meta’s “largest soar” but in coding and agentic work.

There’s substance behind each components of that declare. Muse Spark 1.3 makes important features over final month’s 1.2 launch, notably on long-running agent duties. The model builders can entry now can be one of many strongest price-performance choices close to the highest of unbiased mannequin rankings.

Meta’s strongest Muse Spark 1.3 benchmark outcomes come from its max reasoning configuration. Meta says that model remains to be finishing extra security testing and can arrive “shortly”; the third-party benchmarking agency Synthetic Evaluation says it evaluated max in a restricted associate preview, and presently lists no API supplier in any respect for the configuration.

The model broadly rolling out this week via its Muse Code harness and the Meta Mannequin API makes use of Meta’s beforehand accessible reasoning settings, together with xhigh.

That makes the extra related enterprise query not whether or not Muse Spark 1.3 can attain frontier territory, however how shut the mannequin firms can really deploy at present will get — and at what actual price.

The delivery mannequin is excellent, however not the benchmark chief

Meta does disclose outcomes for each configurations in its underlying analysis report, so this isn’t a case of the corporate hiding the deployable mannequin. However its launch supplies prominently showcase the max variant, and a few of the largest scores belong to that configuration.

For instance, Meta experiences GDPval-AA v2 scores of 1,754 Elo for max versus 1,709 for xhigh, OSWorld 2.0 scores of 66.9 versus 57.2, and JobBench scores of 64.9 versus 61.2.

On some exams the excellence is negligible or reversed: DeepSearchQA is tied at 89.4, whereas xhigh scores 89.2 on Terminal-Bench 2.1 versus max at 88.8.

Synthetic Evaluation scores Muse Spark 1.3 max at 62 on its Intelligence Index and the delivery xhigh model at 61. The latter ties GPT-5.6 Sol max, Grok 4.6 excessive and Claude Opus 5 excessive. However Anthropic nonetheless occupies the highest of the leaderboard: Claude Fable 5.1 reaches 66 at max and 65 at xhigh, whereas Claude Opus 5 reaches 63 at max and xhigh.

In different phrases, Muse Spark 1.3 xhigh is legitimately within the frontier cluster, however it isn’t the mannequin presently setting the frontier.

That’s nonetheless a considerable change from Muse Spark 1.2. VentureBeat’s protection of final month’s launch discovered Meta fielding a reputable coding challenger that however usually trailed Anthropic’s finest mannequin. Muse Spark 1.2 scored 82.9% on Terminal-Bench 2.1 versus Opus 5’s 86.7%, and in addition completed behind Opus on the opposite important coding comparisons Meta introduced.

With 1.3, Meta is now not merely displaying up in that contest. On a number of coding and agentic evaluations, it’s buying and selling wins with OpenAI and Anthropic.

Meta says the underlying mannequin has additionally develop into simpler to function. Muse Spark 1.3 is skilled to take care of a number of workflows in an extended thread, collect context with instruments, detect gaps in its personal plans, ask customers for clarification when obligatory and make sure earlier than consequential actions. In Meta engineers’ inner comparisons, it used roughly 20% fewer instrument calls and 25% fewer tokens than 1.2 throughout coding work.

For enterprises paying for hundreds or tens of millions of agent loops, these behavioral enhancements might matter greater than one other leaderboard level.

‘Virtually too low cost to meter’ doesn’t imply Meta lower its costs

Muse Spark 1.3 didn’t obtain an API value lower. Meta stored Commonplace pricing precisely the place it was for Muse Spark 1.2: $1.25 per million enter tokens, $4.25 per million output tokens and $0.15 per million cached enter tokens.

That makes Zuckerberg’s “nearly too low cost to meter” line much less a press release about decrease token costs than about what Meta believes builders can accomplish with these tokens.

Synthetic Evaluation provides proof for that argument, but in addition a complication. It measures Muse Spark 1.3 xhigh at 235.2 output tokens per second and estimates a price of $0.55 per Intelligence Index process.

At 61 on the Intelligence Index, that offers it the lowest price per process of any presently measured mannequin at that intelligence stage.

Muse Spark 1.2 price solely $0.40 per Synthetic Evaluation process, whereas scoring 57.

Regardless of unchanged per-token pricing, the unbiased benchmark’s price of finishing a median process due to this fact elevated generation-over-generation. Synthetic Evaluation attributes the rise primarily to heavier input-token consumption on agentic evaluations.

That doesn’t straight contradict Meta’s declare of 25% decrease token use: Meta is describing comparisons in its personal coding workflows, whereas Synthetic Evaluation is measuring a broader suite of reasoning and agentic duties.

Nevertheless it illustrates why “low cost” turns into slippery as soon as fashions function as brokers. Token charges, reasoning effort, variety of turns, instrument calls and retries all contribute to the precise price of ending work.

Meta additionally retains its unusually low cost Contributor tier — $0.10 per million enter tokens and $0.20 per million output tokens — in alternate for permission to make use of prompts and completions for coaching.

As VentureBeat famous with Muse Spark 1.2, which may be engaging for prototyping however creates a materially completely different data-governance calculation for enterprises working with proprietary code or delicate inner data.

Wang’s ‘Gemini who?’ lands on an unusually shut comparability

Meta chief AI officer Alexandr Wang was significantly much less certified in celebrating the discharge.

After Synthetic Evaluation posted its Muse Spark outcomes, Wang reposted them on X, including: “i actually hate to say it, however… gemini who? ”

The shade was notably pointed as a result of Google launched Gemini 3.8 Flash on the identical day, pitching it at nearly precisely the identical class of workload: long-horizon software program engineering, autonomous brokers and multi-step skilled reasoning. Google calls 3.8 its finest reasoning and coding Flash mannequin but and says it’s the firm’s third Flash launch in six weeks.

Impartial numbers give Wang one thing to work with, although hardly a knockout.

Synthetic Evaluation offers Muse Spark 1.3 xhigh a 61 Intelligence Index rating at $0.55 per process, in contrast with 59 and $0.58 for Gemini 3.8 Flash at excessive reasoning. Meta due to this fact edges Google on each intelligence and process price at these explicit settings.

Google wins decisively on throughput. Synthetic Evaluation measures Gemini 3.8 Flash excessive at about 305 output tokens per second, versus 235 for Muse Spark — roughly 30% quicker. Gemini additionally has the decrease uncooked API sticker value for now: Google is charging an introductory $0.75 per million enter tokens and $3.75 per million output tokens, in contrast with Meta’s $1.25 and $4.25.

That promotional Google pricing expires December 31, after which it rises to $1.50 per million enter tokens and $7.50 per million output tokens.

The result’s a helpful snapshot of how tight frontier-model economics have develop into. Meta presently wins this unbiased comparability by two Intelligence Index factors and three cents per benchmark process; Google provides considerably greater output throughput and cheaper uncooked tokens throughout its launch promotion.

Wang’s “gemini who?” is enjoyable govt trash speak. For an enterprise architect, the reply is nearer to: Gemini is the quicker choice; Muse is presently the marginally stronger high-effort agent by this unbiased measure.

Meta’s evolving open weights stance

The extra consequential difficulty for some builders might have little to do with at present’s benchmark race.

When Meta launched Muse Code and Muse Spark 1.2 in August, VentureBeat famous how dramatically the corporate had moved away from the open-weight technique that made Llama ubiquitous.

Muse Code and Spark 1.2 have been proprietary, API-served merchandise — a putting posture for the corporate that had spent years arguing that open AI was the trail ahead.

5 days later, Meta modified course once more.

On August 10, it launched the 30-billion-parameter Muse Glimmer below an Apache 2.0 license. Zuckerberg additionally mentioned: “Within the coming weeks, we’re additionally going to open the weights for Muse Spark 1.2.” Reuters individually reported Meta’s plan to launch the Spark 1.2 weights.

Now, Meta has as an alternative shipped Muse Spark 1.3 as one other proprietary mannequin.

That doesn’t but quantity to a damaged promise — “coming weeks” can fairly describe a interval longer than three weeks. However at present’s announcement makes the roadmap much less clear slightly than extra.

Meta’s new put up now not says Muse Spark 1.2. It says its roadmap contains “the Muse Spark open weights launch”, with out figuring out a model, launch date, mannequin measurement or license. Zuckerberg likewise mentioned on X that “Muse Spark open weights releases” are coming quickly.

For groups that standardized on Llama as a result of downloadable weights meant self-hosting, customization and management over inference economics, that ambiguity might matter greater than whether or not Spark gained one other level on a composite benchmark.

Muse Spark 1.3 exhibits that Meta can now iterate proprietary frontier fashions at extraordinary pace. The delivery xhigh configuration is quick, competitively priced and far nearer to the highest of unbiased rankings than its predecessors. The max preview exhibits Meta can push the household a little bit additional when allowed to spend extra reasoning compute.

The subsequent take a look at is completely different: whether or not Meta can convert that tempo right into a roadmap enterprises can really plan round — together with making its finest capabilities broadly deployable and delivering the open-weight Spark mannequin it has already mentioned is coming.