Synthszr Charts — die großen AI-Marken im Wettkampf ums Podium

Llama 3.1-8B

Meta · v3.1 · 8b · 4× · last seen Aug 09, 2026

Momentum

Llama 3.1-8B is an open-weight language model developed by Meta with 8 billion parameters, released on July 23, 2024 as part of the Llama 3.1 model family (8B, 70B, 405B). It is a purely text-based, multilingual autoregressive Transformer model featuring Grouped-Query-Attention, a context window of 128,000 Token, and a training data cutoff of December 2023. The model is licensed under the Llama 3.1 Community License, is freely downloadable, and is offered as an API by numerous third-party providers (e.g., Groq, DeepInfra, AWS Bedrock) at varying prices, as Meta itself does not operate its own paid Inference API.

Momentum trend

11.06.09.09.

Features

Key Benchmark (%)	MMLU (Instruct, 5-shot): 69.4%
Context Window (Tokens)	128,000 tokens
License	Llama 3.1 Community License (allows commercial use with restrictions)
Multimodality	No – text-only model (input: text, output: text)
Platform	Open-weight model, downloadable via Hugging Face/Meta; deployable locally or via third-party APIs (e.g. AWS Bedrock, Groq, DeepInfra)
Price per 1M Tokens	$0.02 (input) / $0.04 (output) per 1M tokens, example OpenRouter; Meta itself does not operate its own paid API
Release Date	July 23, 2024

More products in this category: Open-Source LLMs

Sources (4)

Company Analysis: Meta

As of 08/09/2026

HOLDSynthszr Vote

HOLD is warranted given the combination of (1) strong, data-supported ad momentum (Q2 2026 impressions +14% YoY; price per ad +12% YoY) and constructive Q3 revenue guidance ($61–$64B), versus (2) unusually high and still-rising investment intensity (FY2026 capex $130–$145B; FY2026 expenses $165–$169B) that can constrain margins and free cash flow in the next several quarters. Regulatory headline risk in the EU (DSA preliminary finding on “addictive design”) adds uncertainty around product changes and compliance costs. Net: fundamentals support continued revenue growth, but the risk-adjusted return profile is balanced until investors see clearer evidence that AI capex translates into durable incremental profit/FCF rather than primarily higher cost structure. (sec.gov)

Summary

Meta Platforms, Inc. (META) operates a two-segment model: (1) Family of Apps (FoA)—Facebook, Instagram, WhatsApp, Messenger—monetized primarily via performance advertising; and (2) Reality Labs (RL)—AR/VR hardware and software plus AI-enabled wearables—monetized via device sales and related content/services. Core competencies are large-scale consumer product distribution, ad targeting/measurement, and increasingly AI-driven ranking, creative, and advertiser tooling. Market position remains top-tier in global digital advertising, with durable advantages from scale (multi-billion-user reach), engagement, and a broad advertiser base. In the most recent quarter (Q2 2026, ended June 30, 2026), Meta reported strong ad demand and engagement: ad impressions across FoA rose 14% YoY and average price per ad rose 12% YoY, supporting robust revenue growth. (sec.gov) Competitive pressure (notably short-form video and retail media) persists, but Meta’s AI-driven ad stack and cross-app distribution continue to support share stability. Recent developments (last ~90 days) are dominated by (a) Q2 2026 earnings and (b) elevated AI infrastructure investment. Meta guided Q3 2026 total revenue to $61–$64B (with ~1% FX headwind) and raised/narrowed FY2026 total expense guidance to $165–$169B. (investor.atmeta.com) It also narrowed FY2026 capex (including finance leases) to $130–$145B as it scales AI compute and data centers. (investing.com) Profitability remains strong but more volatile due to investment: the company reported an operating margin of ~31% in Q2 2026 (as presented in its results materials). (investor.atmeta.com) RL continues to be a material drag: RL revenue was $431M in Q2 2026 (+16% YoY), while RL operating losses remained substantial per company filings. (investor.atmeta.com) Valuation/metrics (current price in USD; user requested EUR): META’s trailing P/E is ~23.24x per market data. Converting the share price to EUR requires a live FX rate (not provided here); therefore, this analysis keeps valuation multiples in USD terms while focusing on fundamentals. Outlook (short- to medium-term): Near-term revenue visibility is supported by Q3 guidance and continued ad pricing/impression momentum, but margins and free cash flow are likely to be constrained by the very high capex and expense outlook tied to AI infrastructure. Regulatory risk in the EU remains active; the European Commission issued a preliminary finding (July 10, 2026) alleging “addictive design” breaches under the Digital Services Act for Instagram/Facebook, which could lead to remedies and/or penalties. (digital-strategy.ec.europa.eu)

Key Takeaways

Advertising engine remains the primary value driver: Q2 2026 FoA ad delivery and pricing both grew (impressions +14% YoY; price per ad +12% YoY), supporting strong top-line momentum. (sec.gov)
Near-term revenue guide is constructive: Q3 2026 total revenue guidance of $61–$64B provides a measurable demand signal (with ~1% FX headwind). (investor.atmeta.com)
Investment intensity is the central swing factor: FY2026 capex guidance of $130–$145B and FY2026 expense guidance of $165–$169B imply sustained pressure on margins/FCF versus a lower-investment regime. (investing.com)
Reality Labs remains financially dilutive: Q2 2026 RL revenue was $431M (+16% YoY), but operating losses remain large per filings, limiting consolidated margin expansion. (sec.gov)
EU regulatory overhang is active: the Commission’s July 10, 2026 preliminary DSA finding on “addictive design” increases compliance/remedy risk for engagement mechanics and product design. (digital-strategy.ec.europa.eu)

Action Ideas

BUY

12–18 month accumulation thesis focused on FoA earnings power and AI-enabled ad performance. Data points supporting the thesis include Q2 2026 ad impression and pricing growth (volume +14% YoY; price +12% YoY) and Q3 revenue guidance of $61–$64B, indicating continued demand. The key underwriting assumption is that incremental AI capex/opex sustains product and ad performance improvements enough to keep revenue growth healthy while consolidated margins stabilize after the peak build phase.

Horizon: 18 mo.

HOLD

6–12 month risk-managed hold for investors already positioned, emphasizing valuation discipline and monitoring of investment efficiency. With a trailing P/E around ~23.24x, the stock is not priced as a distressed asset; the near-term debate is whether AI infrastructure spending yields measurable ROI without structurally lowering margins. Maintain exposure while requiring evidence in upcoming quarters that revenue growth and operating margin can coexist with the elevated capex plan.

Horizon: 9 mo.

SELL

3–6 month de-risking thesis for investors with low tolerance for cash-flow volatility. The company’s FY2026 capex guidance ($130–$145B) and raised expense outlook ($165–$169B) increase the probability that near-term FCF remains constrained even if revenue grows, which can be unfavorable in higher-rate or risk-off tape conditions. This action is most appropriate if the portfolio mandate prioritizes near-term cash generation and lower regulatory headline risk.

Horizon: 6 mo.

Contrarian Insights

• Consensus often frames AI capex as purely margin-negative; a contrarian framing is that Meta’s unusually strong ad pricing + volume growth in Q2 2026 suggests AI-driven ranking/creative/measurement improvements are already monetizing, potentially offsetting part of the investment burden sooner than expected (watch ad impressions and price per ad as leading indicators). (sec.gov)
• Reality Labs is frequently treated as a perpetual cash drain; a contrarian angle is that RL’s AI glasses revenue growth (+16% YoY to $431M in Q2 2026) indicates at least one subcategory is scaling, which could gradually improve the segment’s revenue base even if operating losses remain large in the medium term. (investor.atmeta.com)