NewzBench — Which AI Can Predict Tomorrow's News?

📋 Backtest Details — 4 Real-World Events

Geopolitics

Iran War Prediction

Given a week of Iran escalation headlines, predict what happens next.

2026-02-27

📰 Headlines given to models (10)

Trump could attack Iran in days — what's at stake for the oil market (CNBC)
America's most powerful warship heads to Middle East (Fox News)
Iran reportedly preparing for war (Fox News)
Iran strikes likely as Trump seeks maximum leverage (The Hill)
Trump considering limited military strike against Iran (CNBC)
Israelis keep suitcases packed as Trump weighs Iran strike (Fox News)
Iran one week from bomb-grade uranium, protests flare in Tehran (TOI)
Western forces in Erbil relocate ahead of possible US strikes (MEE)
Iran-US talks mediated by Oman begin in Geneva to avert war (Dawn)
Vance says no chance US gets into prolonged Middle East war (Fox News)

✅ What Actually Happened

US and Israel launched massive strikes on Iran. Ayatollah Khamenei killed. Full-scale war began. Oil surged 35%. Middle East airspace closed.

Model Scores

haiku-4.5

3.2

kimi-k2.5

3.0

opus-4.6

3.0

mistral-large-3

2.6

nova-2-lite

2.4

deepseek-v3.2

2.4

glm-4.7

2.2

nova-premier

2.2

qwen3-next-80b

2.2

llama4-maverick

2.0

gpt-oss-120b

2.0

sonnet-4.6

2.0

🔮 View each model's predictions

haiku-4.5 3.2/5

• US military strike or limited air campaign against Iranian nuclear/military facilities occurs (9/10)

• Iran launches retaliatory missile or drone strikes against US/Israeli targets (8/10)

• Oil prices spike 10-20% on Middle East conflict escalation (8/10)

• Diplomatic talks in Geneva collapse or stall without breakthrough (7/10)

kimi-k2.5 3.0/5

• US conducts limited military strikes on Iranian nuclear or military facilities (7/10)

• Iran retaliates with missile attacks on US bases or regional targets (Israel, Gulf states) (6/10)

• Oman-mediated talks produce temporary de-escalation agreement or extension (5/10)

• Oil prices spike 10-20% on supply disruption fears or actual Strait of Hormuz tensions (8/10)

opus-4.6 3.0/5

• The US launches limited precision strikes against Iranian nuclear enrichment facilities (likely Ford (7/10)

• Oman-mediated US-Iran talks in Geneva collapse or are suspended without agreement within 2-4 days, w (8/10)

• Oil prices spike to $95-$110+ per barrel within days as markets price in imminent conflict and poten (8/10)

• Iran activates proxy forces (Hezbollah, Iraqi militias, Houthis) to strike US bases in Iraq/Syria an (7/10)

mistral-large-3 2.6/5

• The U.S. launches a limited military strike on Iranian nuclear or military facilities, targeting enr (8/10)

• Iran retaliates with cyberattacks on U.S. or Israeli infrastructure (e.g., power grids, financial sy (7/10)

• Oil prices spike above $120/barrel as markets price in Middle East supply disruptions, with Saudi Ar (9/10)

• Protests in Tehran and other Iranian cities escalate into violent clashes with security forces, lead (6/10)

nova-2-lite 2.4/5

• Limited US military strike against Iran (8/10)

• Escalation of tensions between Iran and Israel (7/10)

• Continued mediation efforts by Oman to avert war (9/10)

• Flare-up of protests in Tehran (7/10)

deepseek-v3.2 2.4/5

• The United States conducts a limited, targeted military strike against Iranian military or nuclear i (8/10)

• Iran responds asymmetrically, launching missile or drone attacks against US assets or allies in the (7/10)

• Oil prices surge by 10-15% following the US strike and Iranian retaliation, with volatility spiking (8/10)

• The Oman-mediated talks in Geneva break down or are suspended indefinitely following the military ac (7/10)

glm-4.7 2.2/5

• The US launches limited military strikes against Iranian military infrastructure or nuclear faciliti (8/10)

• Iran announces a significant advancement in uranium enrichment levels or officially breaches nuclear (7/10)

• Global oil prices surge above $100 per barrel following the outbreak of hostilities. (9/10)

• Oman-mediated negotiations in Geneva collapse as both the US and Iran refuse to de-escalate. (6/10)

nova-premier 2.2/5

• US conducts limited military strikes on Iranian nuclear facilities (8/10)

• Oil prices surge by 10-15% due to geopolitical tensions (9/10)

• Emergency UN Security Council meeting called to address US-Iran conflict (7/10)

• Protests intensify in Tehran and other major Iranian cities (6/10)

qwen3-next-80b 2.2/5

• The United States launches a limited precision airstrike on Iranian nuclear facilities near Natanz o (9/10)

• Iran retaliates with drone and missile attacks on U.S. military bases in Iraq and/or Israel, trigger (8/10)

• Iran-US talks in Geneva collapse after the U.S. strike, with Iran withdrawing its delegation and acc (7/10)

• Oil prices surge above $95/barrel as markets panic over Middle East supply risks, prompting emergenc (8/10)

llama4-maverick 2.0/5

• US launches a limited military strike against Iran (7/10)

• Iran-US diplomatic talks in Geneva yield a tentative agreement to avert war (6/10)

• Protests in Tehran escalate in response to Iran's nuclear program and potential US strike (8/10)

• Western forces in Erbil and other regional locations take further defensive or preparatory measures (8/10)

gpt-oss-120b 2.0/5

• The United States launches a limited, precision strike on Iranian air defense and nuclear-related fa (8/10)

• Iran retaliates within 48‑72 hours of the U.S. strike by firing a barrage of ballistic missiles and (7/10)

• Oman‑mediated Iran‑U.S. talks in Geneva break down on March 3, 2026, leading both sides to issue sta (6/10)

• Western embassies and contractors in Erbil, Iraq, begin an organized evacuation of non‑essential per (7/10)

sonnet-4.6 2.0/5

• US and Iran reach a preliminary agreement or framework during Oman-mediated Geneva talks, pausing im (6/10)

• Iran announces a temporary freeze or partial rollback of uranium enrichment activities as a negotiat (5/10)

• Limited US or Israeli airstrikes targeting Iranian nuclear or military facilities occur (5/10)

• Iran conducts retaliatory proxy attacks against US assets or bases in Iraq or Syria (6/10)

💡 {ev.get("notable","")}

Technology

Apple Product Launch

Apple leaks and rumors — predict what gets announced.

2026-02-25

📰 Headlines given to models (4)

Apple's Touch-Screen MacBook Pro To Have Dynamic Island (Slashdot)
MacBook Pro with OLED touch screen arriving in the fall (Tom's Hardware)
Apple will start making Mac minis in the US (Engadget)
Apple's autonomous AI SOC (9to5Mac)

✅ What Actually Happened

Apple announced iPhone 17e, MacBook Neo ($599 laptop — surprise new product line), M4 iPad Air, and 7 new products at March event.

Model Scores

haiku-4.5

1.5

llama4-maverick

1.4

nova-premier

1.4

gpt-oss-120b

1.0

kimi-k2.5

1.0

nova-2-lite

1.0

mistral-large-3

1.0

glm-4.7

1.0

qwen3-next-80b

1.0

sonnet-4.6

1.0

opus-4.6

1.0

deepseek-v3.2

1.0

🔮 View each model's predictions

haiku-4.5 1.5/5

• Apple releases official specifications or technical details for the OLED MacBook Pro with Dynamic Is (7/10)

• Apple announces US manufacturing expansion details for Mac mini production facility (6/10)

• Tech analysts or Apple supply chain sources publish deeper specifications about the autonomous AI SO (5/10)

• Apple stock experiences volatility or analyst upgrades based on these product announcements (6/10)

llama4-maverick 1.4/5

• Apple announces plans to integrate more AI into their ecosystem (7/10)

• Details about the manufacturing process for the US-made Mac mini are revealed (6/10)

• Apple's stock price reacts positively to the news of US-based Mac mini production (8/10)

• Rumors or leaks about the specifications of the new touch-screen MacBook Pro surface online (9/10)

nova-premier 1.4/5

• Apple to announce pre-order dates for the new Touch-Screen MacBook Pro (8/10)

• Further details emerge on Apple's plans for US-based Mac mini production (7/10)

• Leaked benchmarks or specifications of Apple's new autonomous AI SOC appear online (6/10)

• Industry analysts release reports comparing the new MacBook Pro to existing models (7/10)

gpt-oss-120b 1.0/5

• Apple will hold a special product‑focused livestream on February 28, 2026, unveiling a MacBook Pro p (9/10)

• Apple will file a trademark registration for "Dynamic Island" covering macOS devices, which will bec (8/10)

• Apple will release a detailed press release on March 2, 2026, confirming that the first U.S.-made Ma (7/10)

• Apple will submit an SEC filing on March 3, 2026, disclosing a new internal project code‑named "Proj (6/10)

kimi-k2.5 1.0/5

• Apple officially announces Touch-Screen MacBook Pro with Dynamic Island at a product event or press (8/10)

• Apple confirms US manufacturing expansion for Mac mini with specific facility location (7/10)

• Apple reveals details of custom AI chip/SOC for autonomous systems or vehicles (6/10)

• Apple stock experiences notable volatility based on mixed reception to touch-screen Mac strategy (5/10)

nova-2-lite 1.0/5

• Apple announces official release timeline for the new MacBook Pro with OLED touch screen and Dynamic (9/10)

• Apple reveals details about its autonomous AI System on a Chip (SOC) at a technical session or devel (7/10)

• Apple shares production plans for US-made Mac Mini devices, including possible factory locations and (8/10)

• Industry analysts publish early performance benchmarks or reviews of the new MacBook Pro's Dynamic I (6/10)

mistral-large-3 1.0/5

• Apple announces a March 5 'Peek Performance' event to unveil the new MacBook Pro with OLED touch scr (8/10)

• Leaked benchmarks of Apple's autonomous AI SOC surface online, showing a 30-40% performance boost ov (7/10)

• Foxconn or another Apple supplier confirms plans to expand US production of the Mac mini, with a new (6/10)

• A developer discovers references to 'Dynamic Island' in macOS 15 beta code, confirming cross-platfor (7/10)

glm-4.7 1.0/5

• Apple announces a supply chain partnership for US-based Mac mini manufacturing, revealing the specif (8/10)

• Developers discover code references to 'Dynamic Island' support in macOS betas ahead of the new MacB (9/10)

• Analysts release reports estimating the cost impact of OLED panels on the upcoming MacBook Pro prici (7/10)

• Apple publishes a research paper or job posting related to the specific capabilities of their new 'a (6/10)

qwen3-next-80b 1.0/5

• Apple announces prototype of MacBook Pro with Dynamic Island and OLED touch screen at a developer ev (9/10)

• Apple reveals first internal details of its autonomous AI SOC, including architecture and power effi (8/10)

• Apple opens a small-scale Mac mini assembly line in Austin, Texas, as part of its US manufacturing i (7/10)

• Apple files a patent for touch-sensitive bezel integration with Dynamic Island on MacBook Pro (6/10)

sonnet-4.6 1.0/5

• Apple officially announces or leaks additional details about the Touch-Screen MacBook Pro's Dynamic (7/10)

• Apple confirms a specific US manufacturing facility or partner (likely in Texas or Arizona) for Mac (6/10)

• Analyst reports or supply chain leaks emerge providing more technical specifications for the autonom (6/10)

• Competitors (likely Microsoft Surface or Dell) respond to Apple's touch-screen MacBook Pro announcem (5/10)

opus-4.6 1.0/5

• Apple officially announces expanded US manufacturing plans for Mac mini at a dedicated press event o (7/10)

• Tech analysts and teardown channels publish detailed speculation and leaked specs about Apple's auto (8/10)

• Apple suppliers (Samsung Display, LG Display) confirm or leak details about OLED touch-screen panel (6/10)

• Competing laptop manufacturers (Dell, Lenovo, Microsoft) issue statements or tease their own touch-s (5/10)

deepseek-v3.2 1.0/5

• Apple issues official press release detailing the Touch-Screen MacBook Pro's Dynamic Island feature (8/10)

• Apple announces specific US manufacturing location and initial production timeline for the Mac mini. (7/10)

• Technical analysts and chip experts publish deep dives into the architecture and capabilities of App (6/10)

• Increased speculation and sourcing about which Mac models will be first to use the new 'autonomous A (5/10)

💡 {ev.get("notable","")}

Technology

OpenAI Pentagon Controversy

Pentagon pressures AI companies on military contracts.

2026-02-26

📰 Headlines given to models (5)

Hegseth to meet Anthropic CEO as Pentagon threatens banishment (Axios)
US Defense Secretary summons Anthropic CEO for tough talks over military AI (Straits Times)
The US military will reportedly use Musk's Grok AI in classified systems (Engadget)
Anthropic CEO Says AI Company 'Cannot In Good Conscience Accede' To Pentagon (Slashdot)
Anthropic vs. the Pentagon (Slate)

✅ What Actually Happened

OpenAI made controversial Pentagon surveillance deal. Altman admitted it was 'opportunistic and sloppy'. ChatGPT uninstalls surged 295%. Anthropic refused. Major backlash.

Model Scores

sonnet-4.6

2.8

opus-4.6

2.6

haiku-4.5

1.8

deepseek-v3.2

1.8

nova-premier

1.6

nova-2-lite

1.4

kimi-k2.5

1.4

glm-4.7

1.3

mistral-large-3

1.2

llama4-maverick

1.0

gpt-oss-120b

1.0

qwen3-next-80b

1.0

🔮 View each model's predictions

sonnet-4.6 2.8/5

• Anthropic and Pentagon reach a formal impasse or Anthropic publicly reaffirms refusal to provide Cla (8/10)

• Pentagon formally announces expanded contract or partnership with xAI to deploy Grok in additional c (7/10)

• Other major AI companies (OpenAI, Google DeepMind, or Palantir) publicly signal willingness to fill (7/10)

• Congressional hearings or formal legislative discussion is announced regarding AI companies' obligat (6/10)

opus-4.6 2.6/5

• Anthropic formally declines or places significant restrictions on a Pentagon military AI contract, i (8/10)

• The Pentagon announces expanded partnership with xAI (Grok) for classified military systems, positio (7/10)

• Congressional Republicans call for investigation or legislative action against Anthropic, framing th (7/10)

• Other major AI companies (OpenAI, Google DeepMind, or Meta) are publicly asked about or pressured to (6/10)

haiku-4.5 1.8/5

• Hegseth-Anthropic CEO meeting occurs, resulting in public statement about disagreements over militar (9/10)

• Anthropic issues formal policy statement reaffirming refusal to provide AI for classified military a (8/10)

• Pentagon announces or accelerates plans to expand Grok AI integration into classified systems as alt (7/10)

• Congressional inquiry or statement issued regarding AI companies' obligations to support national de (6/10)

deepseek-v3.2 1.8/5

• Anthropic CEO Dario Amodei publicly refuses to sign a formal agreement with the Pentagon, citing the (9/10)

• The Pentagon announces a provisional ban on using Anthropic's Claude models in any military or intel (8/10)

• Elon Musk or xAI announces a finalized contract with the U.S. Department of Defense for the integrat (7/10)

• A significant public and industry debate erupts, with other AI CEOs (from OpenAI, Google, etc.) bein (8/10)

nova-premier 1.6/5

• Anthropic CEO and Pentagon officials will hold a press conference addressing the ongoing dispute. (8/10)

• US Congress schedules hearings to discuss the ethical implications of AI in military applications. (7/10)

• Tech industry leaders release a joint statement on AI ethics in response to the Anthropic-Pentagon s (6/10)

• The Pentagon announces a temporary halt in integrating new AI systems until ethical guidelines are r (5/10)

nova-2-lite 1.4/5

• Anthropic CEO formally refuses Pentagon's AI deployment demands during meeting with Defense Secretar (9/10)

• Pentagon issues formal ultimatum or threat of sanctions to Anthropic for non-compliance (8/10)

• Musk's Grok AI begins limited integration into classified US Defense systems (7/10)

• Congressional hearings announced to investigate ethical AI use in military applications (6/10)

kimi-k2.5 1.4/5

• Anthropic CEO Dario Amodei publicly clarifies or partially walks back refusal to work with Pentagon, (7/10)

• Congressional hearings or bipartisan inquiry announced on military AI procurement standards and vend (6/10)

• Elon Musk/XAI issues statement defending Grok's security credentials and attacking Anthropic's relia (8/10)

• Tech industry coalition or open letter emerges supporting Anthropic's stance, with signatories from (5/10)

glm-4.7 1.3/5

• The Pentagon officially announces a contract to integrate Elon Musk's Grok AI into classified milita (8/10)

• Anthropic publicly releases a detailed 'Responsible AI Policy' explicitly defining the boundaries of (7/10)

• Defense Secretary Hegseth issues a formal warning or policy directive requiring AI contractors to co (6/10)

mistral-large-3 1.2/5

• Anthropic announces a formal policy rejecting military contracts for AI development, citing ethical (8/10)

• Pentagon accelerates integration of Grok AI into classified systems, bypassing Anthropic for key pro (9/10)

• US lawmakers introduce bipartisan legislation to regulate AI-military collaborations, with specific (7/10)

• Anthropic CEO meets with EU officials to explore partnerships and regulatory alignment for civilian (6/10)

llama4-maverick 1.0/5

• Anthropic CEO meets with Pentagon officials to discuss AI usage (8/10)

• Pentagon announces decision on Anthropic's AI usage in military systems (7/10)

• US military begins integrating Musk's Grok AI into classified systems (6/10)

• Hegseth publicly discusses meeting with Anthropic CEO (5/10)

gpt-oss-120b 1.0/5

• Deputy Secretary of Defense Kathleen Hegseth meets with Anthropic CEO Dario Amodei at the Pentagon f (9/10)

• The Pentagon issues a formal directive suspending all existing Anthropic contracts and places the co (8/10)

• The U.S. Senate Armed Services Committee convenes a closed‑door briefing on AI in defense, featuring (7/10)

• The Department of Defense announces an accelerated rollout of Elon Musk’s Grok AI into classified we (6/10)

qwen3-next-80b 1.0/5

• Anthropic publicly announces suspension of all military AI contracts and withdraws from Pentagon dis (9/10)

• Pentagon officially approves deployment of Musk's Grok AI in at least two classified defense systems (8/10)

• Hegseth and Anthropic CEO hold a televised, adversarial hearing before Congress on AI ethics and nat (7/10)

• Major tech partners of Anthropic (e.g., Amazon, Google) issue public statements distancing themselve (6/10)

💡 {ev.get("notable","")}

Business

Oil Surge & Market Crash

Iran war just started — predict market impact.

2026-02-28

📰 Headlines given to models (5)

Israel targets Khamenei, top leaders in bid to bring down Iran's regime (Axios)
Trump's unprovoked attack on Iran risks terror, turmoil and spread of nuclear weapons (Independent)
US and Israel attack Iran. Bitcoin and oil offer a glimpse at how markets react (MarketWatch)
Oil prices surge as Middle East conflict escalates (CNBC)
Defense stocks soar as Iran conflict widens (MarketWatch)

✅ What Actually Happened

Oil surged 35% in one week (biggest gain since 1983). Dow futures sank 1000+ points. Gas prices spiked. Defense-tech stocks boomed. Treasury market worst weekly rout since 'liberation day'.

Model Scores

nova-2-lite

2.8

nova-premier

2.8

llama4-maverick

2.2

opus-4.6

2.2

haiku-4.5

1.8

deepseek-v3.2

1.8

gpt-oss-120b

1.4

kimi-k2.5

1.4

glm-4.7

1.4

qwen3-next-80b

1.2

sonnet-4.6

1.2

mistral-large-3

🔮 View each model's predictions

nova-2-lite 2.8/5

• Intensified airstrikes and military operations in Iran by US and Israeli forces targeting key milita (9/10)

• Global oil prices continue to spike due to supply disruptions from the Middle East conflict, affecti (10/10)

• US and its allies impose new economic sanctions on Iran aimed at crippling the regime’s financial an (8/10)

• Protests and civil unrest grow inside Iran as the regime faces internal pressure following attacks o (7/10)

nova-premier 2.8/5

• Emergency UN Security Council meeting called to address the escalating conflict between the US, Isra (9/10)

• Further increase in global oil prices due to sustained tensions in the Middle East. (8/10)

• Iran-backed militias in the region launch retaliatory attacks against US or Israeli interests. (7/10)

• Major stock markets experience volatility as investors react to geopolitical tensions. (8/10)

llama4-maverick 2.2/5

• US and Israeli retaliation against Iran-backed militias in the region (8/10)

• Global economic instability due to rising oil prices (9/10)

• Iranian retaliation against US and Israeli targets (7/10)

• Emergency meeting of the UN Security Council to address the crisis (6/10)

opus-4.6 2.2/5

• Iran or Iranian proxies (Hezbollah, Houthis, Iraqi militias) launch retaliatory strikes against US m (9/10)

• Oil prices spike above $110/barrel, with possible brief spikes toward $120+, as markets price in pot (8/10)

• UN Security Council emergency session convenes, with China and Russia vetoing or blocking US-backed (8/10)

• US military announces deployment of additional carrier strike groups and troop reinforcements to the (9/10)

haiku-4.5 1.8/5

• Oil prices exceed $100-120 per barrel as markets price in sustained supply disruption from Iran conf (9/10)

• Iran launches retaliatory missile or drone strikes against US/Israeli targets or regional allies (8/10)

• Defense contractor stock indices (RTN, LMT, NOC, GD) rise 3-8% as investors anticipate sustained mil (8/10)

• Bitcoin and cryptocurrencies surge 5-15% as investors seek inflation hedges and assets outside tradi (7/10)

deepseek-v3.2 1.8/5

• Iran launches retaliatory missile and drone strikes against Israeli military and strategic targets. (9/10)

• The United States publicly confirms its involvement in the strikes and announces a significant deplo (8/10)

• Global oil prices breach $120 per barrel as shipping disruptions are reported in the Strait of Hormu (8/10)

• Emergency meeting of the UN Security Council is called, resulting in a deadlock and failed resolutio (7/10)

gpt-oss-120b 1.4/5

• Iran launches a coordinated missile and drone barrage against U.S. and Israeli assets in the Gulf re (8/10)

• The United Nations Security Council convenes an emergency session to condemn the escalation and call (7/10)

• U.S. Treasury Department imposes a new round of secondary sanctions targeting Iranian oil export inf (7/10)

• Global oil prices (Brent) breach $130 per barrel, while WTI climbs above $125, driven by fears of su (9/10)

kimi-k2.5 1.4/5

• Iran launches retaliatory missile strikes against Israeli military and civilian targets (8/10)

• US deploys additional carrier strike group and air defense assets to Middle East (9/10)

• Saudi Arabia and UAE issue joint statement calling for immediate ceasefire while secretly coordinati (7/10)

• Bitcoin volatility exceeds 40% daily swings as investors flee to gold and Swiss franc (6/10)

glm-4.7 1.4/5

• Crude oil prices (Brent/WTI) will exceed $110 per barrel as supply disruptions from the Persian Gulf (9/10)

• Iran will launch retaliatory ballistic missile strikes against US military bases in the Middle East (8/10)

• Major stock indices (S&P 500, Nasdaq) will correct by at least 5% as volatility spikes. (7/10)

• Bitcoin will experience extreme volatility (swings of 10% or more) but trend upward as a 'safe haven (6/10)

qwen3-next-80b 1.2/5

• Iran launches retaliatory missile strikes against Israeli military bases and U.S. facilities in the (9/10)

• Oil prices rise above $110/barrel as Saudi Arabia and UAE temporarily reduce exports amid fears of S (8/10)

• Bitcoin surges past $85,000 as global investors flock to it as a 'digital safe haven' amid fears of (7/10)

• U.S. Congress holds emergency session to debate authorization of expanded military operations in the (8/10)

sonnet-4.6 1.2/5

• Oil prices reach $120-130 per barrel as Strait of Hormuz faces disruption threats or partial blockad (8/10)

• Iran launches ballistic missile or drone retaliatory strikes against US bases in Iraq, Qatar, or aga (8/10)

• Emergency UN Security Council session convenes, with Russia and China vetoing Western resolutions wh (7/10)

• Bitcoin surges past $120,000 as investors treat it as a geopolitical hedge alongside gold breaking $ (6/10)

💡 {ev.get("notable","")}

Which AI can predict
tomorrow's news?

Only 2 out of 8 models predicted US military strikes on Iran

Iran War Prediction

Apple Product Launch

OpenAI Pentagon Controversy

Oil Surge & Market Crash

Which AI can predicttomorrow's news?

Only 2 out of 8 models predicted US military strikes on Iran

Iran War Prediction

Apple Product Launch

OpenAI Pentagon Controversy

Oil Surge & Market Crash

Which AI can predict
tomorrow's news?