Will GigaChat release an open-weights model with ≥100B parameters by the end of 2026?
Russia / UkraineOne-OffIA3mManifold MarketsSem KYCDados de resolução verificadosRazoavelmente calibradoPrevisão da comunidade atualNão 54.8%
Previsores
1
Tipo de pergunta
binary
Metodologia
Play-money forecasting platform
Tipo de fonte
Previsão
Dados do mercado
Atualizado há 12 horas
Desatualizado29/10/25, 13:25 UTC1/01/27, 23:59 UTC
Tendências
Mercados Relacionados
Perguntas relacionadas para dar contexto, não uma comparação de contratos equivalentes.
Will Anthropic release an open-weights model in 2026?
Manifold MarketsWill an Open Weights Model be #1 on the Artificial Analysis Intelligence Index in 2026
Manifold MarketsOpen Weights text-to-image model with Elo ≥ 1200 on Artificial Analysis by EOY 2026?
Manifold MarketsWill Anthropic sign the Open Weights and American AI Leadership letter?
KalshiTrump, Putin, and Zelensky seen together before 2027?
PolymarketNATO x Russia military clash by which date?
PolymarketRegras
In June 2025, the GigaChat team released three open-source models based on the Mixture of Experts (MoE) architecture with 20 billion total parameters (3.3 billion active).
- GigaChat is developed by SaluteDevices, a subsidiary of Sberbank, one of Russia's largest financial institutions and a leading player in Russian AI development.
- While the team has released smaller open-weights models, they maintain proprietary versions (GigaChat Pro and GigaChat MAX) for which no architectural details or weights have been disclosed.
- The progression to releasing significantly larger open-weights models would provide signal towards several important developments in Russia's AI ecosystem:
- Computational capacity: Training models above 100B parameters requires substantial GPU clusters and energy infrastructure.
- Delivering a ≥100B judged worthwhile to make open-weights would suggest access to a reasonable scale of AI infrastructure
Mercados abertos
Russia / UkrainePOLYMARKETRussia Elections: United Russia Wins Every Region?
- Yes+5.0pt69.3%
- No-5.0pt30.7%
- Volume total
- € 630 mil
- Volume 24h
- € 145,7 mil
POLYMARKETNATO x Russia military clash by which date?
- December 31+0.5pt29%
- October 3116%
- December 31, 2025-0.1pt0%
+3 mais resultados
POLYMARKETPutin out as President of Russia by which date?
- June 30, 2027-1.0pt13%
- December5%
- July0%
+2 mais resultados
- Volume total
- € 21,8 mi
POLYMARKETWill Russia capture Orikhiv by which date?
- December 31-3.5pt33%
- September 30-1.9pt2%
- Volume total
- € 292 mil
- Volume 24h
- € 36,4 mil
POLYMARKETWill Russia enter Dobropillia by which date?
- December 31+6.5pt90%
- October 31+12.5pt73%
- September 30+7.0pt35%
+6 mais resultados
POLYMARKETNobel Peace Prize Winner 2026
- Sudan's Emergency Response Rooms-3.4pt14%
- Save the Children-2.9pt9%
- Yulia Navalnaya8%
+28 mais resultados
- Which parties will win at least one seat in the next Russian State Duma election?
- How many seats will New People (NL) win in the next Russian legislative election?
- Russia Election: 2nd Place Party List in Moscow?
- How many seats will the Liberal Democratic Party of Russia (LDPR) win in the next Russian legislative election?
- Russia Election: United Russia (ER) Vote Share?
- Russia x Ukraine ceasefire agreement by which date?
- Will Russia announce a new forced mobilization by which date?
- Will Ukraine recapture Crimean territory by which date?
Perguntas frequentes
- Quais são as probabilidades atuais de Will GigaChat release an open-weights model with ≥100B parameters by the end of 2026??
- A Manifold Markets cota No em 55% em 20 de setembro de 2026.
- Quando Will GigaChat release an open-weights model with ≥100B parameters by the end of 2026? é resolvido?
- Will GigaChat release an open-weights model with ≥100B parameters by the end of 2026? deverá ser resolvido em 1 de janeiro de 2027.
- Como Will GigaChat release an open-weights model with ≥100B parameters by the end of 2026? é liquidado?
- A Manifold Markets liquida este mercado da seguinte forma: In June 2025, the GigaChat team released three open-source models based on the Mixture of Experts (MoE) architecture with 20 billion total parameters (3.3 billion active).