Anthropic Reveals Two Claude Models, Marshmallow Offers Better Conversational Experience than Opus 5
By: x.com|2026/08/24 02:12:59
0
Share
Prefer us on Google
Developers have discovered two early access Claude models from Anthropic, codenamed claude-marshmallow-eap and claude-melon-eap. Tester feedback indicates that Marshmallow performs overall better than Melon, but neither has reached the level of Fable 5. Additionally, some testers reported that Marshmallow's conversational experience surpasses that of Opus 5. Previously, Anthropic has named early models using a "food name + EAP" convention. In early July, Claude Honeycomb EAP briefly appeared on Cursor before being removed; on July 24, Anthropic officially released Opus 5, but the company has not confirmed a direct relationship between Honeycomb and Opus 5.
-- Price
--
--
--
This content is provided for general informational purposes only and doesn't constitute financial, investment, legal, or tax advice. Any events, rewards, online promotions, or related information mentioned herein should not be considered a recommendation, solicitation, or invitation to purchase, sell, trade, or otherwise deal in any crypto assets. Crypto assets are highly volatile and may result in loss. The availability of WEEX services, products, and related events may vary by region. You are responsible for ensuring that your participation is in accordance with applicable local laws and regulations.
You may also like
Fake Claude App 101MB Attempts to Steal Cryptocurrency Wallets
A malicious campaign to extract cryptocurrency wallets and account information through fake desktop installation files has been confirmed at the time of Anthropic's release of the new Claude model.
Fake Claude App Spreads RevStealer, Targeting 50 Crypto Wallets
Fable 5 Expenditure at 11.4%, Changes in AI Adoption
Anthropic's top AI model, Claude Fable 5, accounted for 11.4% of corporate spending in July. Companies are increasingly favoring lower-cost AI models.
Corporate Clients Opt for Cheaper Models Instead of Claude Fable 5
Qwen 3.8-27B Surpasses Muse Glimmer, Approaches Claude Opus 4.6
GPT-5.6 Sol Costs 13% Less Than Kimi K3
The belief that Chinese AI models are cheaper than American ones has been challenged in actual cost comparisons. While the token price is lower for Chinese models, the number of tokens used for answer generation and quality also play a crucial role.
Claude Sonnet 5 to Transition to Standard Pricing on September 1
The price of Anthropic's Claude Sonnet 5 API is set to transition to standard pricing on September 1.
Cases of AI Agent Accessing External Infrastructure
Cases of AI agents accessing actual external infrastructure during internal cybersecurity assessments have been repeatedly confirmed.
Alibaba internally announced a complete ban on Claude, which will take effect on July 10
According to reports from ZhDongxi, Alibaba announced today an全面 ban on the Claude series products under Anthropic. All employees are required to uninstall Agent products, including multiple series models such as Sonnet, Opus, Fable, and Claude Code. This ban will officially take effect on July 10. ...
Zcash vulnerability researcher will audit Monero and plans to expand to more privacy coin projects
According to CoinDesk, security engineer Taylor Hornby, who discovered a serious vulnerability in Zcash using the Anthropic Opus 4.8 AI model, stated that Monero (XMR) has been added to the audit queue, and more privacy coin projects will undergo security reviews in the future.Hornby previously disc...
B.AI subsidy increase: Limited time opening of 1:1 equivalent recharge rebate mechanism
According to official news, B.AI launched a significant recharge benefits upgrade event on May 3, temporarily opening a 1:1 equivalent recharge rebate mechanism. Users can enjoy an equivalent amount of credits upon recharge, achieving an ultra-low cost of 50% for accessing cutting-edge models like G...
Microsoft enables GPT to collaborate with Claude, outperforming all competitors
Microsoft announced on Monday the launch of two new features for Copilot Researcher—Critique and Council—combining OpenAI's GPT with Anthropic's Claude for the same research task.Critique adopts a sequential collaboration model: GPT is responsible for planning the research, retrieving information, a...
OpenClaw has designated Venice AI as the recommended model provider, and the VVV token has risen over 70% in the past 7 days
Founder of Venice AI, Erik Voorhees, posted on the X platform that the OpenClaw documentation shows Venice has been set as the recommended model provider, with the default being venice/llama-3.3-70b and "best overall" being venice/claude-opus-45.Market data shows that Venice Token (VVV) has risen by...
Bitcoin Price Ahead of Jobs Report: Has Crypto Already Priced In the Fed's Next Move?
Bitcoin holds near $77.7K ahead of the September 4 jobs report. See how Fed rate expectations, NFP forecasts and key BTC levels could shape the next move.
Arthur Hayes: Bitcoin Could Reach One Million Dollars by 2030, but Ethereum is My Top Choice
Isabel Schnabel to Leave ECB for IMF Position Before 2027
The potential early departure of Isabel Schnabel from the European Central Bank to take a position at the International Monetary Fund would open a politically sensitive succession in Frankfurt, just as inflation in the eurozone remains above target and the official advocates for a restrictive moneta...
NDV: Bitcoin, the Core Asset in the Era of Dollar and Bitcoin Overissuance
Crypto: The Fake GTA 6 That Drains Your Wallet
A fake GTA 6 fan site sells a "leaked" copy and loads a wallet drainer capable of emptying a wallet across seven blockchains.
Bitcoin, Ethereum, Solana: Who is really the most decentralized?
Robinhood put stocks on a permissionless blockchain- memecoin traders are stress-testing what happens next
Memecoin traders are turning Robinhood’s tokenized stocks into liquidity for a new wave of 24/7 crypto markets.
Bitcoin: Buy High and Sell Low, Strategy Justifies and Explains Its Choices
Strategy claims to have sold 7,000 bitcoins between $60,000 and $65,000 before resuming purchases above $80,000.
Strong Yen Pressures Nikkei and Exposes Bank of Japan's Dilemma
Bitcoin Golden Cross Approaches: The 12 Previous Instances That Call for Caution
Bitcoin is approaching its golden cross, a signal with an uneven track record. This time, the expansion of USDT supply lends credibility to the bullish reading.
Standard Chartered brings institutional Bitcoin, Ether trading to UAE
43% of Indonesian Fintechs Have Turned a Profit, IDX Ready to Support IPOs
The Indonesian fintech industry is entering a more mature growth phase, with 43% of members reporting profits and 86% adopting AI.
Tether’s USDT0 launches on Stellar with cross-chain liquidity
Interest Rate Hike Probability Rises from 37% to 67% in a Week, New York Fed President: Long-Term Bond Yields Reflect Economic Strength
Hyperliquid Strategies Increases Equity Financing Limit to $2.5 Billion (Additional Submission to SEC)
10 Trillion Euros in Dormant Savings: What the EU Wants to Do with Your Money
"Lazy" savings: von der Leyen wants to mobilize 10 trillion euros of European deposits. Requisition? No. What the speech really says.
XRP Ledger tested by BIS researchers for data checks
Fake Claude App 101MB Attempts to Steal Cryptocurrency Wallets
A malicious campaign to extract cryptocurrency wallets and account information through fake desktop installation files has been confirmed at the time of Anthropic's release of the new Claude model.
Fake Claude App Spreads RevStealer, Targeting 50 Crypto Wallets
Fable 5 Expenditure at 11.4%, Changes in AI Adoption
Anthropic's top AI model, Claude Fable 5, accounted for 11.4% of corporate spending in July. Companies are increasingly favoring lower-cost AI models.
Corporate Clients Opt for Cheaper Models Instead of Claude Fable 5
Qwen 3.8-27B Surpasses Muse Glimmer, Approaches Claude Opus 4.6
GPT-5.6 Sol Costs 13% Less Than Kimi K3
The belief that Chinese AI models are cheaper than American ones has been challenged in actual cost comparisons. While the token price is lower for Chinese models, the number of tokens used for answer generation and quality also play a crucial role.
...
Customer Support:@weikecs
Business Cooperation:@weikecs
Quant Trading & MM:bd@weex.com
VIP Program:support@weex.com
