Pick Atria Dawn Preview for best-in-class among tested models on several of shanghai ai lab's own benchmarks, including browsecomp (92.5, ahead of gpt-5.6 sol's 92.2) and deepsearchqa (96.0) or 744b-parameter moe built on a glm-5.2 base, released under a fully open mit license. Pick Muse Spark 1.1 for scaled tool use — 88.1 on mcp atlas, ahead of opus 4.8 and gpt-5.5 (vendor-reported) or subagent orchestration — trained to run as a main agent or a subagent that escalates when stuck. Choose Atria Dawn Preview if you need self-hosting or data privacy; Muse Spark 1.1 if you want a managed API.
Atria Dawn Preview (Shanghai AI Laboratory, China) and Muse Spark 1.1 (Meta, US) line up two different AI ecosystems against each other — a comparison that is as much about cost philosophy and openness as raw capability. Atria Dawn Preview is a free, MIT-licensed 744B open-weight model from Shanghai AI Lab that beat GPT-5.6 Sol on the BrowseComp benchmark — released quietly, weights before paper. Muse Spark 1.1 is meta's first paid, closed-weight frontier model — class-leading agentic tool use at a quarter of rivals' price, but it trails on coding. They diverge most on price, context window and open vs. closed weights — each quantified below from the models' real specs.
Key differences
Cost model: Atria Dawn Preview ships open weights you can self-host (hardware cost only, no per-token fee), while Muse Spark 1.1 is API-metered at $1.25/$4.25 per 1M tokens. Your choice depends on whether you want zero marginal cost at the price of running infrastructure.
Context window: Muse Spark 1.1 holds 4× more — 1M (~1,573 pages) vs 256K tokens (~393 pages). But effective recall usually fades long before the advertised ceiling, so the bigger number only helps if the model reasons over it.
Recency: Atria Dawn Preview is the newer model by about 2 months (released September 11, 2026), usually meaning fresher training data and capabilities.
Ecosystem: this is a China-vs-US matchup — they differ in pricing philosophy, data-residency options, and tooling ecosystems, not only benchmarks.
Specifications
Spec
Atria Dawn Preview
Muse Spark 1.1
Provider
Shanghai AI Laboratory (China)
Meta (US)
Released
September 11, 2026
July 9, 2026
Context window
256K tokens (~393 pages)
1M (~1,573 pages)
Price (in/out)
Open weight (self-host / free)
$1.25/$4.25 per 1M tokens
Open weight?
Yes — self-hostable
No — API only
Modalities
text
text, image, video, code
SWE-Bench Verified
Not published
Not published
MRCR v2 @ 1M
Not published
54.1%
Who wins what
Best-in-class among tested models on several of Shanghai AI Lab's own benchmarks, including BrowseComp (92.5, ahead of GPT-5.6 Sol's 92.2) and DeepSearchQA (96.0): Atria Dawn Preview — A free, MIT-licensed 744B open-weight model from Shanghai AI Lab that beat GPT-5.6 Sol on the BrowseComp benchmark — released quietly, weights before paper — and its weights are open while Muse Spark 1.1 is API-only.
744B-parameter MoE built on a GLM-5.2 base, released under a fully open MIT license: Atria Dawn Preview — Open weights make this possible at all — Muse Spark 1.1 is API-only, so it cannot leave the vendor's servers.
Free to self-host — no API pricing, run entirely on your own hardware: Atria Dawn Preview — A free, MIT-licensed 744B open-weight model from Shanghai AI Lab that beat GPT-5.6 Sol on the BrowseComp benchmark — released quietly, weights before paper — and it is the newer of the two.
Scaled tool use — 88.1 on MCP Atlas, ahead of Opus 4.8 and GPT-5.5 (vendor-reported): Muse Spark 1.1 — Atria Dawn Preview is comparatively weak here — context window is reported inconsistently across trackers — most list 256K tokens, at least one lists 1M; unconfirmed which is accurate
Subagent orchestration — trained to run as a main agent or a subagent that escalates when stuck: Muse Spark 1.1 — Meta's first paid, closed-weight frontier model — class-leading agentic tool use at a quarter of rivals' price, but it trails on coding — and it carries the larger 1M context.
Professional agentic work — 54.7 on JobBench, a wide margin over rivals (vendor-reported): Muse Spark 1.1 — Muse Spark 1.1 lists professional agentic work — 54.7 on JobBench, a wide margin over rivals (vendor-reported) among its strengths; Atria Dawn Preview does not.
Lowest cost at scale: Atria Dawn Preview — Its weights are open, so at volume you pay for your own hardware instead of Muse Spark 1.1's $1.25/$4.25 per 1M tokens.
Largest single-prompt input: Muse Spark 1.1 — Its 1M window is about 4× larger than Atria Dawn Preview's 256K tokens, fitting roughly 1,573 pages in one prompt.
Which should you pick?
A cost-sensitive startup shipping high volume: Atria Dawn Preview — At Open weight (self-host / free) it undercuts Muse Spark 1.1, and on millions of tokens that margin decides the monthly bill.
Someone analysing very long documents or codebases: Muse Spark 1.1 — Larger 1M window fits more in one prompt.
A team with data-privacy or self-hosting needs: Atria Dawn Preview — Open weights let you run it on your own hardware; Muse Spark 1.1 is API-only.
Anyone whose priority is best-in-class among tested models on several of shanghai ai lab's own benchmarks, including browsecomp (92.5, ahead of gpt-5.6 sol's 92.2) and deepsearchqa (96.0): Atria Dawn Preview — It is specifically built for that.
Anyone whose priority is scaled tool use — 88.1 on mcp atlas, ahead of opus 4.8 and gpt-5.5 (vendor-reported): Muse Spark 1.1 — That is its strongest area.
An enterprise with regional data-residency rules: Muse Spark 1.1 or Atria Dawn Preview — Origin (China vs US) affects where data is processed and which compliance regime applies — check the provider's terms for your region.
Atria Dawn Preview: where it fits
A free, MIT-licensed 744B open-weight model from Shanghai AI Lab that beat GPT-5.6 Sol on the BrowseComp benchmark — released quietly, weights before paper. Released September 11, 2026 by Shanghai AI Laboratory, it is built for best-in-class among tested models on several of Shanghai AI Lab's own benchmarks, including BrowseComp (92.5, ahead of GPT-5.6 Sol's 92.2) and DeepSearchQA (96.0), 744B-parameter MoE built on a GLM-5.2 base, released under a fully open MIT license, and free to self-host — no API pricing, run entirely on your own hardware.
Its trade-offs are real: context window is reported inconsistently across trackers — most list 256K tokens, at least one lists 1M; unconfirmed which is accurate, released quietly, with weights and an API posted to GitHub before Shanghai AI Lab's own technical report went up on arXiv days later, and a lab research preview rather than a commercial product — support and update cadence are unclear. As an open-weight model, its running cost is your own hardware rather than a per-token fee.
Muse Spark 1.1: where it fits
Meta's first paid, closed-weight frontier model — class-leading agentic tool use at a quarter of rivals' price, but it trails on coding. Released July 9, 2026 by Meta, it is built for scaled tool use — 88.1 on MCP Atlas, ahead of Opus 4.8 and GPT-5.5 (vendor-reported), subagent orchestration — trained to run as a main agent or a subagent that escalates when stuck, professional agentic work — 54.7 on JobBench, a wide margin over rivals (vendor-reported), and managing its own context: it compacts the 1M window mid-run instead of relying on external windowing.
Its trade-offs: not the coding leader its launch framing implied — Meta's own report concedes it trails Opus 4.8 and GPT-5.5 on every coding benchmark, the 1M window oversells its recall: 54.1 on MRCR v2 at 1M against GPT-5.5's 74.0, closed weights end the free, self-hostable Llama path — this is the first model Meta has charged for, and uS-only public preview behind a waitlist, and every benchmark is vendor-reported with no third-party replication. At $1.25 in / $4.25 out per million tokens, it sits in the mid price band.
The bottom line for this matchup
The defining split here is open vs. closed. Atria Dawn Preview gives you weights you control — self-host it, fine-tune it, keep data in-house, pay only for hardware. Muse Spark 1.1 gives you a managed, always-updated API with no infrastructure to run. Teams with GPUs, privacy requirements, or huge volume often favour the open model; teams that want zero ops and the latest capabilities favour the closed one. Capability is close enough that this operational question, not the benchmark, usually decides it.
Frequently asked questions
Is Atria Dawn Preview or Muse Spark 1.1 better for coding?
Public SWE-Bench figures are not available for either model, so the honest test is your own repository — run an identical real bug through both. By design, Atria Dawn Preview leans toward best-in-class among tested models on several of shanghai ai lab's own benchmarks, including browsecomp (92.5, ahead of gpt-5.6 sol's 92.2) and deepsearchqa (96.0) while Muse Spark 1.1 leans toward scaled tool use — 88.1 on mcp atlas, ahead of opus 4.8 and gpt-5.5 (vendor-reported), and that positioning usually predicts which feels better on your codebase.
Which is cheaper, Atria Dawn Preview or Muse Spark 1.1?
Atria Dawn Preview is open-weight, so self-hosting means no per-token fee (you pay for hardware instead), while Muse Spark 1.1 is API-metered at $1.25/$4.25 per 1M tokens. For most teams without GPUs, the API model is cheaper to start; at very high volume, self-hosting can win.
Which has the bigger context window?
Muse Spark 1.1 — 1M vs 256K tokens, about 4× larger. Useful only if the model actually reasons over the full window, which not all do.
Can I use both Atria Dawn Preview and Muse Spark 1.1 together?
Yes — a multi-model platform like LumiChats gives you Atria Dawn Preview, Muse Spark 1.1 and 40+ others under one ₹69/day pass (about $1/day), so you can draft with one and cross-check with the other instead of buying two subscriptions.
Which is newer, Atria Dawn Preview or Muse Spark 1.1?
Atria Dawn Preview — released September 11, 2026, about 2 months after Muse Spark 1.1.
Atria Dawn Preview vs Muse Spark 1.1
Shanghai AI Laboratory · China | Meta · US · Updated June 2026
Quick verdict
Pick Atria Dawn Preview for best-in-class among tested models on several of shanghai ai lab's own benchmarks, including browsecomp (92.5, ahead of gpt-5.6 sol's 92.2) and deepsearchqa (96.0) or 744b-parameter moe built on a glm-5.2 base, released under a fully open mit license. Pick Muse Spark 1.1 for scaled tool use — 88.1 on mcp atlas, ahead of opus 4.8 and gpt-5.5 (vendor-reported) or subagent orchestration — trained to run as a main agent or a subagent that escalates when stuck. Choose Atria Dawn Preview if you need self-hosting or data privacy; Muse Spark 1.1 if you want a managed API.
Atria Dawn Preview (Shanghai AI Laboratory, China) and Muse Spark 1.1 (Meta, US) line up two different AI ecosystems against each other — a comparison that is as much about cost philosophy and openness as raw capability. Atria Dawn Preview is a free, MIT-licensed 744B open-weight model from Shanghai AI Lab that beat GPT-5.6 Sol on the BrowseComp benchmark — released quietly, weights before paper. Muse Spark 1.1 is meta's first paid, closed-weight frontier model — class-leading agentic tool use at a quarter of rivals' price, but it trails on coding. They diverge most on price, context window and open vs. closed weights — each quantified below from the models' real specs.
Key differences at a glance
▸Cost model: Atria Dawn Preview ships open weights you can self-host (hardware cost only, no per-token fee), while Muse Spark 1.1 is API-metered at $1.25/$4.25 per 1M tokens. Your choice depends on whether you want zero marginal cost at the price of running infrastructure.
▸Context window: Muse Spark 1.1 holds 4× more — 1M (~1,573 pages) vs 256K tokens (~393 pages). But effective recall usually fades long before the advertised ceiling, so the bigger number only helps if the model reasons over it.
▸Recency: Atria Dawn Preview is the newer model by about 2 months (released September 11, 2026), usually meaning fresher training data and capabilities.
▸Ecosystem: this is a China-vs-US matchup — they differ in pricing philosophy, data-residency options, and tooling ecosystems, not only benchmarks.
Side-by-side specs
Spec
Atria Dawn Preview
Muse Spark 1.1
Provider
Shanghai AI Laboratory (China)
Meta (US)
Released
September 11, 2026
July 9, 2026
Context window
256K tokens (~393 pages)
1M (~1,573 pages)
Price (in/out)
Open weight (self-host / free)
$1.25/$4.25 per 1M tokens
Open weight?
Yes — self-hostable
No — API only
Modalities
text
text, image, video, code
SWE-Bench Verified
Not published
Not published
MRCR v2 @ 1M
Not published
54.1%
Who wins what
Best-in-class among tested models on several of Shanghai AI Lab's own benchmarks, including BrowseComp (92.5, ahead of GPT-5.6 Sol's 92.2) and DeepSearchQA (96.0)
Atria Dawn Preview
A free, MIT-licensed 744B open-weight model from Shanghai AI Lab that beat GPT-5.6 Sol on the BrowseComp benchmark — released quietly, weights before paper — and its weights are open while Muse Spark 1.1 is API-only.
744B-parameter MoE built on a GLM-5.2 base, released under a fully open MIT license
Atria Dawn Preview
Open weights make this possible at all — Muse Spark 1.1 is API-only, so it cannot leave the vendor's servers.
Free to self-host — no API pricing, run entirely on your own hardware
Atria Dawn Preview
A free, MIT-licensed 744B open-weight model from Shanghai AI Lab that beat GPT-5.6 Sol on the BrowseComp benchmark — released quietly, weights before paper — and it is the newer of the two.
Scaled tool use — 88.1 on MCP Atlas, ahead of Opus 4.8 and GPT-5.5 (vendor-reported)
Muse Spark 1.1
Atria Dawn Preview is comparatively weak here — context window is reported inconsistently across trackers — most list 256K tokens, at least one lists 1M; unconfirmed which is accurate
Subagent orchestration — trained to run as a main agent or a subagent that escalates when stuck
Muse Spark 1.1
Meta's first paid, closed-weight frontier model — class-leading agentic tool use at a quarter of rivals' price, but it trails on coding — and it carries the larger 1M context.
Professional agentic work — 54.7 on JobBench, a wide margin over rivals (vendor-reported)
Muse Spark 1.1
Muse Spark 1.1 lists professional agentic work — 54.7 on JobBench, a wide margin over rivals (vendor-reported) among its strengths; Atria Dawn Preview does not.
Lowest cost at scale
Atria Dawn Preview
Its weights are open, so at volume you pay for your own hardware instead of Muse Spark 1.1's $1.25/$4.25 per 1M tokens.
Largest single-prompt input
Muse Spark 1.1
Its 1M window is about 4× larger than Atria Dawn Preview's 256K tokens, fitting roughly 1,573 pages in one prompt.
Which should you pick?
A cost-sensitive startup shipping high volume
→ Atria Dawn Preview
At Open weight (self-host / free) it undercuts Muse Spark 1.1, and on millions of tokens that margin decides the monthly bill.
Someone analysing very long documents or codebases
→ Muse Spark 1.1
Larger 1M window fits more in one prompt.
A team with data-privacy or self-hosting needs
→ Atria Dawn Preview
Open weights let you run it on your own hardware; Muse Spark 1.1 is API-only.
Anyone whose priority is best-in-class among tested models on several of shanghai ai lab's own benchmarks, including browsecomp (92.5, ahead of gpt-5.6 sol's 92.2) and deepsearchqa (96.0)
→ Atria Dawn Preview
It is specifically built for that.
Anyone whose priority is scaled tool use — 88.1 on mcp atlas, ahead of opus 4.8 and gpt-5.5 (vendor-reported)
→ Muse Spark 1.1
That is its strongest area.
An enterprise with regional data-residency rules
→ Muse Spark 1.1 or Atria Dawn Preview
Origin (China vs US) affects where data is processed and which compliance regime applies — check the provider's terms for your region.
Atria Dawn Preview: where it fits
A free, MIT-licensed 744B open-weight model from Shanghai AI Lab that beat GPT-5.6 Sol on the BrowseComp benchmark — released quietly, weights before paper. Released September 11, 2026 by Shanghai AI Laboratory, it is built for best-in-class among tested models on several of Shanghai AI Lab's own benchmarks, including BrowseComp (92.5, ahead of GPT-5.6 Sol's 92.2) and DeepSearchQA (96.0), 744B-parameter MoE built on a GLM-5.2 base, released under a fully open MIT license, and free to self-host — no API pricing, run entirely on your own hardware.
Its trade-offs are real: context window is reported inconsistently across trackers — most list 256K tokens, at least one lists 1M; unconfirmed which is accurate, released quietly, with weights and an API posted to GitHub before Shanghai AI Lab's own technical report went up on arXiv days later, and a lab research preview rather than a commercial product — support and update cadence are unclear. As an open-weight model, its running cost is your own hardware rather than a per-token fee.
Muse Spark 1.1: where it fits
Meta's first paid, closed-weight frontier model — class-leading agentic tool use at a quarter of rivals' price, but it trails on coding. Released July 9, 2026 by Meta, it is built for scaled tool use — 88.1 on MCP Atlas, ahead of Opus 4.8 and GPT-5.5 (vendor-reported), subagent orchestration — trained to run as a main agent or a subagent that escalates when stuck, professional agentic work — 54.7 on JobBench, a wide margin over rivals (vendor-reported), and managing its own context: it compacts the 1M window mid-run instead of relying on external windowing.
Its trade-offs: not the coding leader its launch framing implied — Meta's own report concedes it trails Opus 4.8 and GPT-5.5 on every coding benchmark, the 1M window oversells its recall: 54.1 on MRCR v2 at 1M against GPT-5.5's 74.0, closed weights end the free, self-hostable Llama path — this is the first model Meta has charged for, and uS-only public preview behind a waitlist, and every benchmark is vendor-reported with no third-party replication. At $1.25 in / $4.25 out per million tokens, it sits in the mid price band.
The bottom line for this matchup
The defining split here is open vs. closed. Atria Dawn Preview gives you weights you control — self-host it, fine-tune it, keep data in-house, pay only for hardware. Muse Spark 1.1 gives you a managed, always-updated API with no infrastructure to run. Teams with GPUs, privacy requirements, or huge volume often favour the open model; teams that want zero ops and the latest capabilities favour the closed one. Capability is close enough that this operational question, not the benchmark, usually decides it.
Want both Atria Dawn Preview and Muse Spark 1.1 without two subscriptions? LumiChats gives you these plus 40+ models under one ₹69/day pass (about $1/day) — draft with one, cross-check with the other.
Is Atria Dawn Preview or Muse Spark 1.1 better for coding?
Public SWE-Bench figures are not available for either model, so the honest test is your own repository — run an identical real bug through both. By design, Atria Dawn Preview leans toward best-in-class among tested models on several of shanghai ai lab's own benchmarks, including browsecomp (92.5, ahead of gpt-5.6 sol's 92.2) and deepsearchqa (96.0) while Muse Spark 1.1 leans toward scaled tool use — 88.1 on mcp atlas, ahead of opus 4.8 and gpt-5.5 (vendor-reported), and that positioning usually predicts which feels better on your codebase.
Which is cheaper, Atria Dawn Preview or Muse Spark 1.1?
Atria Dawn Preview is open-weight, so self-hosting means no per-token fee (you pay for hardware instead), while Muse Spark 1.1 is API-metered at $1.25/$4.25 per 1M tokens. For most teams without GPUs, the API model is cheaper to start; at very high volume, self-hosting can win.
Which has the bigger context window?
Muse Spark 1.1 — 1M vs 256K tokens, about 4× larger. Useful only if the model actually reasons over the full window, which not all do.
Can I use both Atria Dawn Preview and Muse Spark 1.1 together?
Yes — a multi-model platform like LumiChats gives you Atria Dawn Preview, Muse Spark 1.1 and 40+ others under one ₹69/day pass (about $1/day), so you can draft with one and cross-check with the other instead of buying two subscriptions.
Which is newer, Atria Dawn Preview or Muse Spark 1.1?
Atria Dawn Preview — released September 11, 2026, about 2 months after Muse Spark 1.1.
Specifications and benchmarks reflect publicly reported figures as of June 2026 and may change as providers release updates. Always verify on your own workload.